Employee Engagement

11 downloads 0 Views 2MB Size Report
Apr 15, 2009 - Academy of Management Journal, 37, 1379-. 1391. DiLalla, L. F. (2000). Structural equation modeling: Uses and issues. In Tinsley, H. &. Brown ...
Employee Engagement: The development of a three dimensional model of engagement; and an exploration of its relationship with affective leader behaviours.

by Jonnie Catherine De Lacy BA (Psychology) PG Dip (Organisational Psychology)

A Research Paper Submitted in Partial Fulfilment of the Requirements for the

Master Of Business (Research)

School of Management Queensland University of Technology 2009

ii

School of Management Queensland University of Technology

Author:

De Lacy, Jonnie C.

Title:

The development of a three dimensional model of engagement; and an exploration of its relationship with affective leader behaviours.

Degree:

Master of Business (Research)

Research Adviser:

Bernd Irmer, Ph.D.

Month/Year:

April, 2009

Number of Pages:

186

Style Manual Used: American Psychological Association, 5th edition

iii

ABSTRACT This study was designed to examine affective leader behaviours, and their impact on cognitive, affective and behavioural engagement. Researchers (e.g., Cropanzano & Mitchell, 2005; Moorman et al., 1998) have called for more research to be directed toward modelling and testing sets of relationships which better approximate the complexity associated with contemporary organisational experience. This research has attempted to do this by clarifying and defining the construct of engagement, and then by examining how each of the engagement dimensions are impacted by affective leader behaviours. Specifically, a model was tested that identifies leader behaviour antecedents of cognitive, affective and behavioural engagement. Data was collected from five public-sector organisations. Structural equation modelling was used to identify the relationships between the engagement dimensions and leader behaviours. The results suggested that affective leader behaviours had a substantial direct impact on cognitive engagement, which in turn influenced affective engagement, which then influenced intent to stay and extra-role performance. The results indicated a directional process for engagement, but particularly highlighted the significant impact of affective leader behaviours as an antecedent to engagement. In general terms, the findings will provide a platform from which to develop a robust measure of engagement, and will be helpful to human resource practitioners interested in understanding the directional process of engagement and the importance of affective leadership as an antecedent to engagement.

iv

Statement of Original Authorship The work contained in this thesis has not been previously submitted to meet requirements for an award at this or any other higher education institution. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made.

Signature :

Date : 15 April, 2009.

v

Acknowledgments I would like to thank Dr Chrys Gunesakara, Dr Amanda Gudmundsson and Dr Bernd Irmer for their help and guidance at various stages of this project, as well as their sheer niceness and approachability. I would particularly like to make mention of the latter, Dr Irmer for his exceptional generosity in terms of time, guidance, accessibility, patience and support – particularly through the long dark hours of statistical endeavour. I would also like to acknowledge the support and patience of my family: husband Brett, and sons; Sam, Joe and Ben. They dealt with the inevitable periods of neglect and crankiness with unfailing good humour, irreverence, and importantly provided a much needed sense of perspective when I most needed it!

vi

TABLE OF CONTENTS List of Tables ........................................................................................................... x List of Figures ........................................................................................................ xi Chapter I: Introduction .......................................................................................... 12 Employee Engagement in the Public Sector ..................................................... 17 Public Sector Reform ........................................................................................ 18 High Engagement Leadership in the Public Sector ........................................... 20 The Role of Engaging Leadership ..................................................................... 23 Purpose of the Research .................................................................................... 25 Contribution to Scholarship .............................................................................. 26 Overview of Thesis Structure ............................................................................ 27 Conclusion ......................................................................................................... 28

Chapter II: Literature Review ............................................................................... 30 Why Engagement matters ................................................................................. 31 What is Engagement? ........................................................................................ 38 A new Construct or a new blend of old wines? ............................... 40 Cognitive Engagement ..................................................................... 46 Affective Engagement ...................................................................... 50 Behavioural Engagement ................................................................. 54 Definition of Engagement. ............................................................... 56 The Antecedents (or drivers) of Engagement. .................................................. 58 The role of Engaging Leadership ...................................................................... 63 Direct Leaders as the conduits of Cognitive Engagement ............... 64

vii

How Management Practitioners view Leadership and Engagement. ..................................................................................... 65 Academic Research on Leadership and Engagement. ..................... 68 The role of Affect in Leadership and Engagement .......................... 69 Transformational Leadership ........................................................... 72 Scope, Summary and Hypotheses ..................................................................... 77 Engaging Leader Behaviours ........................................................... 78 Cognitive Engagement ..................................................................... 79 Affective Engagement...................................................................... 79 Behavioural Engagement. ................................................................ 82 Research Model................................................................................................. 83 Hypotheses ........................................................................................................ 84 H1: Affective Leader Behaviours .................................................... 84 H2: Cognitive Engagement .............................................................. 85 H3: Supervisory Commitment ......................................................... 85 H4: Affective Commitment ............................................................. 85 Mediated Relationships ..................................................................................... 86 H5: Mediating Effects of Cognitive Engagement ............................ 87 H6: Mediating Effects of Supervisory Commitment ....................... 87 H7: Mediating Effects of Affective Commitment ........................... 87 Summary ........................................................................................................... 88

Chapter III: Method and Results ........................................................................... 90 Method .................................................................................................................. 91 Sample ............................................................................................................... 91 Participants ........................................................................................................ 92 Procedure .......................................................................................................... 93 Measures ........................................................................................................... 95

viii

Leader Behaviours (TLQ) ................................................................ 96 Cognitive Engagement ..................................................................... 98 Goal and Process Clarity .................................................................. 98 Communication Satisfaction ............................................................ 98 Connection (Line of Sight).............................................................. 99 Affective Engagement. ..................................................................... 99 Behavioural Engagement ............................................................... 101 Results ............................................................................................................. 102 Data Screening ................................................................................................ 102 Measurement Model ........................................................................................ 104 Higher Order Cognitive Engagement Scale ................................... 104 Structural Model .............................................................................................. 112 Summary of Direct and Mediated Relationships ............................................ 117 Direct Relationships ....................................................................... 117 Mediated Relationships .................................................................. 118

Chapter IV: Discussion ....................................................................................... 123 Discussion of Hypotheses ............................................................................... 125 Leader Behaviours ......................................................................... 126 The Role of Cognitive Engagement ............................................... 127 The Role of Supervisory Commitment .......................................... 127 The role of Affective Commitment ................................................ 128 Pathways of Engagement Model .................................................................... 128 Pathway 1 – Social Emotional Support Behaviours – Through Engagement .................................................................................... 129 Pathway 2 – Leader Behaviours to Supervisory Commitment. ..... 131 Comments on the Direction of the Model ...................................... 133

ix

Contribution to the Literature on Engagement................................................ 138 Implications of Research for the Public Sector ............................................... 141 ‘Employing’ Employee Engagement in the Public Sector Context ................. 142 Limitations of the Research ............................................................................ 146 Future Directions in Engagement Research .................................................... 148 Conclusion ...................................................................................................... 150

References ........................................................................................................... 152

x

List of Tables Table 1: Summary of Measures used ................................................................................ 96  Table 2: Internal consistency and correlation coefficients for TLQ (public) .................... 97  Table 3: Items removed ................................................................................................... 106  Table 4: Confirmatory Factor Item Loadings ................................................................. 106  Table 5: Confirmatory Factor Analyses of Model Fit Indicies ....................................... 110  Table 6: Means, Standard Deviations (SD), Inter-correlations and Internal Consistency Alphas for the Variables .......................................................................................... 111  Table 7: Direct and indirect effects and 95% confidence intervals for the final model .. 120 

xi

List of Figures

Figure 1: Macey and Schneider, (2008) model of employee engagement ........................ 42  Figure 2: Three dimensional construct of engagement ..................................................... 46  Figure 3: Dimensions of engagement and key associated constructs ............................ 566  Figure 4: Engagement sequence...................................................................................... 788  Figure 5: Proposed model of employee engagement ...................................................... 844  Figure 6: Baron and Kenny's mediation model .............................................................. 866  Figure 7: Participating Organisations.............................................................................. 922  Figure 8: Fully mediated model .................................................................................. 11313  Figure 9: Final model .................................................................................................. 11616  Figure 10: Final engagement model ............................................................................ 12929 

12

Chapter I: Introduction In recent years the term “employee engagement” has taken a central role in discussions on organisational efficacy. This is due in part to its touted impact on employee outcomes, describing a deeper level of involvement, passion and enthusiasm for work than other similar constructs such as organisational commitment and job satisfaction. This thesis examines how ‘employee engagement’, a relatively new and inadequately researched construct, impacts on how much effort an employee is willing to invest in a work role (referred to as discretionary effort), and her/his intention to remain with an organisation (employee retention). It will also examine the important role of supportive leadership on employee engagement. There has been a proliferation of practitioner literature supporting the notion that engaged employees positively impact on their organisation’s performance and sustainability (TowersPerrin-ISR, 2006). Organisations that have traditionally relied upon financial measures such as profitability, revenue, and expenditure control are increasingly considering the so-called "soft", human-oriented measures such as employee attitudes, turnover, and levels of engagement as important predictors of organisational efficacy and sustainable competitive advantage. This change in focus has evolved in the context of the transformation of work witnessed over the last two decades. The high number of corporate restructurings, technological advances, and competitive pressures that began in the early 1980s has impacted on current work practices by revolutionising the structure of organisations, the terms and conditions of work and redefining the relationship between employer and employee (Welbourne, 2007). For businesses striving to improve productivity in a global environment where intense competition is constraining profit margins, and

13

where attracting and retaining the right people is becoming increasingly difficult, the concept of employee engagement takes on new appeal. Employee retention, in particular, is the current key focus of many industry sectors in virtually all developed countries. To pick up a newspaper is to be assailed by claims of “crisis” because of skills shortages and an inability to stem the turnover tide (Welbourne, 2007). Practitioner commentary overwhelmingly asserts that the ability to both attract and retain talented employees, is the most critical factor to influence the effectiveness of organisations and will be a key issue for the foreseeable future (Bates, 2004; Coopers, 2004; Frank, Finnegan, & Taylor, 2004). The consequences for businesses that fail to pay attention to these issues are significant and are measured in terms of the cost of employee turnover and other indirect costs such as an inability to expand production, lost productivity, erosion of morale and falling profitability (Frank et al., 2004; Towers Perrin, 2004; 2006). At the macro level, labour market data indicates that there will be significant reductions in the number of new employees entering the workplace, placing likely constraints on the ability of firms to expand production, or maintain existing levels of productivity (Bates, 2004). Several reasons for this exist, including: the aging of the workforce and lower birth rates, and the subsequent inability to fill the vacancies left by retiring baby boomers; the impacts of globalisation (leading to increased competitiveness in the quest for talent); and changing expectations of workers. Moreover, these problems are being experienced in most industrialised countries, not just Australia – a worldwide labour market shortage appears imminent and protracted, in the developed world at least (Gantz Wiley Research, 2004).

14

The tightening labour market has serious financial consequences for business, as skilled workers have increasing options to move between jobs. In the USA, voluntary job turnover (the term used when employees choose to seek employment elsewhere) is estimated to cost the economy $5 trillion a year (Journal of Business strategy, 2003). Turnover costs include the costs of advertising, recruitment and selection and associated costs such as travel and relocation, orientation and induction. They also include significant indirect costs such as lost productivity, lost sales opportunities, erosion of morale, and the time and labour spent up-skilling new employees (Corporate Leadership Council, 2004). Another contributor to the job turnover mix is changing worker expectations. Research conducted in industrialised countries indicates widespread disengagement and dissatisfaction in the workplace, with a significant percentage of workers preparing to leave their current employer once labour market shortages take hold (Bates, 2004; PricewaterhouseCoopers, 2002). In an environment of increasing opportunities, employees are able to shop their skills for better conditions and higher pay and jobs which satisfy their expectations on a range of levels (Bates, 2004). This is particularly acute in high job skill areas and the professions (Frank et al., 2004; Jesuthasan, 2003) . The impact of globalisation, skills shortages and widespread dissatisfaction in the workplace has given rise to a change in power relations between employers and employees. Employees have increasing number of choices available to them – if they do not like how they are being treated by one firm, they can simply pick up a job elsewhere. Further, these choices are no longer local, as job seekers are finding demand for their skills internationally as well as in other locations within in their

15

own countries. As a consequence, there has been increased pressure on organisations to rethink the way in which they can engage and retain employees (Nelson, 1997; Human Resources Institute, 2004). The change in employer-employee power relations has created an environment where business has become increasingly interested in issues such as employee loyalty, trust, and commitment (Tsui & Wu, 2005). Moreover, businesses are seeking support and advice on how best to build organisation cultures and climates where these intrinsic employee beliefs and attitudes can develop and flourish (Masson, Royal, Agnew & Fine, 2008). Employee engagement offers a suitable intervention, having the capacity to “turn on” cognitively and behaviourally absent employees. For example, practitioner research has shown that organisations in which employees experience high levels of engagement, significantly outperform their peers (Towers Perrin, 2004). ISR, an employee research and consulting firm, conducted a global engagement study of more than 664,000 employees which showed that a “well-substantiated relationship exists between employee engagement—the extent to which employees are committed, believe in the values of the company, feel pride in working for their employer, and are motivated to go the extra mile—and business results”, the latter being defined as operating income, net income, and earnings per share (EPS) over a 12 month period (TowersPerrin-ISR, 2006, p.8). Further research suggests that firms that routinely practice employee engagement techniques have higher productivity and profitability than comparative firms in the same industry that do not apply these interventions (Bates, 2004; TowersPerrin-ISR, 2006). In addition, research indicates that employee engagement offers more than just reduced employee turnover and consequential improved cost control.

16

Workplaces that focus on employee engagement appear to produce higher levels of customer satisfaction as well, leading to improved customer retention and increased sales (Frank et al., 2004). The behaviours learned and practiced in engaged workplaces appear to be transferable to customer relationships – the extent of which was summarised by Hewitt Associates, a HR consulting company, who claim that they “have established a conclusive, compelling relationship between engagement and profitability through higher productivity, sales, customer satisfaction, and employee retention” (Hewitt Associates LLC, 2001, p. 1). Highly engaged workforces can produce both cost efficiencies and revenue growth – somewhat of a “holy-grail” in business management terms. In this context, it is easy to see why a growing number of employers are attracted to the notion of employee engagement. Picking up on this interest, employee engagement is heavily marketed by human resource (HR) consulting firms and practitioners. However, the methods and interventions used to create highly engaged workplaces lack a strong empirical base. As a consequence, employee engagement is open to allegations that it is more about ‘spin than substance’. As will be discussed in subsequent chapters, academic researchers are only just starting to inform practice in the field. Accordingly, despite a surge in interest in improving engagement from the business community, HR practitioners and researchers still disagree about what employee engagement is, how to go about getting it, and what it looks like when it is achieved. Additionally, with all the attention given to reported levels of employee engagement and the subsequent impact on organisational efficacy, there are few if any indicators on what a realistic

17

level of engagement should be for employees overall and for various subgroups of workers (Welbourne, 2007). The challenge for researchers is in clarifying the competing and inconsistent interpretations of the meaning of the construct, and in providing suitable and agreed boundaries around it, such that it can measured and interpreted both in a qualitative and quantitative sense. Without such clarity, those interested in impacting on the construct in the workplace risk implementing strategies that have little or no impact, discrediting it as a meaningful and legitimate subject for management attention (Macey & Schneider, 2008). Suffice to say at this stage, if employee engagement is going to be a credible field of research, and one that can inform practice, a clear definition of the construct needs to be provided. Dealing with this issue of “definition” and its applicability to the workplace will be a key focus of the next chapter. Employee Engagement in the Public Sector This study will focus specifically on employee engagement in the public sector. The preceding section discussed employee engagement in workplaces generally, with a number of references to engagement in the private sector context (e.g., examples that focused on issues of profitability and cost control – items that are generally considered more relevant to the private corporate sector). At face value, the distinction between the public and private sectors could create the perception that the two environments are quite different, and as a result, employee engagement practices in the private sector are not applicable to the public sector, and vice versa.

18

This study adopts the position that the literature on employee engagement is applicable to both the private and public sectors due to two fundamental reasons: first, the public sector reform processes of the past twenty years has been focused on making the public sector more like the private sector; and second, the issue of leadership, and more specifically, leadership styles that use high employee engagement tactics are not private or public sector biased. In fact, employee engagement in the public sector has potentially greater currency than in the private sector, as will be argued below. Public Sector Reform There has been radical and ongoing reform of the public sector in most Western countries over the past twenty years, as governments have tried to deal with the challenges of increased political involvement in decision-making, technological change, globalisation and international competition (Colley, 2001; Halligan, 1997; Halligan & Power, 1992). The changes in public sector management practices, grouped under the banner of “New Public Management” gained momentum in the 1980s, and reached widespread appeal with virtually all of the government administrations in Australia in the early-to-mid 1990s (Gould-Williams, 2003). The changes represented a paradigm shift in the way public sector resources were organised, including Australian governments (Colley, 2001; Halligan 1997). Previously, the public sector was organised in a military-style fashion, with rigid rules and procedures dictating work practices and operating systems. For the public sector in Australia, traditional bureaucratic management practices, inherited from its British Westminster background, defined the type of work to be done and the process by which it was undertaken. Inflexible pay-scales and job classification processes

19

were inextricably linked to an internal hierarchical chain of command, where promotion through the ranks was based on seniority. The traditional public sector was seen as “internally focused”, with a closed labour market, and with permanent tenure being a hallmark of the system – a job in the public sector was a “job-for-life.” By the 1980s, these practices were viewed as inefficient and incapable of meeting the challenges that confronted the public sector at the time, particularly in relation to cost control, innovation and service standards (Colley, 2001; Denhardt & Denhardt, 2000). The New Public Management framework that swept through Australian governments, and other western democracies, had a number of defining traits including: “hands on” professional management of the public sector by a cadre of contracted Senior Executive Officers; clear agency and individual accountability for outputs and, at times, outcomes; a focus on disaggregating large monolithic government structures into smaller units that were expected to reach defined performance targets; contracting arrangements that split “purchasers” from “providers” and introduced open competition with the private sector in some circumstances; tighter budget and resource management systems (often expressed as the expectation to “do more with less”); and a general application of – sometimes idealised - private sector management practices to public administration systems (Pollitt, 1993). The reforms of the 1980s and 1990s witnessed an opening up of public sector employment, with jobs advertised broadly (including private sector-like recruitment advertisements in the press), and recruitment and selection processes oriented towards open merit, as opposed to seniority, or length of service. Applicants from

20

outside the public sector were actively encouraged to pursue public sector employment, particularly for contracted or senior level positions (Colley, 2001). Also, the contracting of senior positions enabled political representatives to have a greater say in appointment processes, ensuring that appointments shared the political objectives of the government-of-the-day. The system that evolved was described as “Washminster”, in a play-on-words with the traditional bureaucratic structures associated with Westminster government and the system that operates in the United States (ie. Washington). It was evident that New Public Management brought with it elements of the American system, where the heads of public sector departments were political appointments that changed with the election of a new government (Colley, 2001; Weller & Wood, 1999). High engagement leadership in the public sector Irrespective of the political motivation for greater involvement in senior executive appointment, the message for the senior executives in the new order of management was for them to be more “private sector-like”, adopting leadership and management principles which were hitherto more applicable to their private sector counterparts (Colley, 2001; Palmer & Dunford, 2001). The New Public Management has proven to be quite resilient with the reform process continuing on through the early part of the twenty-first century (Halligan, 2005). The blending of leadership and management principles across the private and public sectors has given rise to a number of areas of common ground. For instance, the modern public sector, like its private counterpart, is deemed to be suffering from

21

rising levels of disengagement, dissatisfaction and employee turnover (Albrecht, 2005). The notion of employee engagement has particular resonance in the public sector as even with the reforms that have taken place, the sector still has a number of restrictions in relation to the salaries and benefits it can provide to its employees. This leaves issues such as leadership styles and engagement strategies as potent potential interventions to differentiate the public sector and provide a tangible competitive advantage over the private sector. Developing competitive advantage is important in the new “open” public sector, given the tight labour market shortages of the mid 2000s (Wenham, 2005). In response to these acknowledged new challenges (Albrecht, 2005; Lawler, 2008; Trottier, Van Wart, & Wang, 2008), new models of public sector leadership have emerged. Senior executive competency standards suggest that “leadership in the work environment” that “facilitates workforce effectiveness” and “building and maintaining relationships” are critical to public sector management. Within broader competency standards, “fostering mutual trust and respect in the work environment” has been explicitly acknowledged as a key element of public sector leadership effectiveness (PSMO, 1998.) Essentially, the public service is being forced to examine the factors that will enable it to retain staff and thereby, knowledge and service bases in their organisations (Hughes, 1994). As a result, there is a growing demand for leaders and leadership able to carry out these tasks, and to see through fundamental processes of change (Lawler, 2008). In the public and not for profit (NFP) sectors, leadership, rather than management, has been identified as the key requirement in making the

22

progressive changes necessary for ‘modernisation’ and effectiveness in the 21st century. It is acknowledged, that leaders, rather than mangers, have the skills to transform organisations by their vision for the future and the articulation of their vision so that others are empowered and enabled to take responsibility for achieving it (Alimo-Metcalfe & Lawler, 2001). Rodgers et al. (2003) describe the increased interest in leadership in the public sector, and the attendant expectations of leaders, as constituting ‘a veritable rush to leadership’ (p. 2). Certainly, the focus of much of the public sector leadership literature individualises leadership, that is, it suggests leadership as residing with a single individual who is expected to influence other individuals/groups (Lawler, 2008). Albrecht (2005) notes that in much of the public sector literature, related engagement constructs such as employee commitment and increased motivation have been touted as the most effective means of achieving organisational performance and that ‘transformational leadership’ is seen as the means of developing it. Put another way, competitive advantage in the public sector can be gained from recruiting, selecting, and retaining leaders who will optimise opportunities of achieving organisational success (Albrecht, 2005). Given that the importance of leadership effectiveness has been acknowledged in the modern public sector, an understanding of the impacts of public sector leadership needs to be clearly articulated and grounded firmly in theory (Trottier, Van Wart & Wang, 2008). Systematic research is needed to determine the specific leader behaviours that contribute to engagement. There still appears to be little, if any, attention by leadership researchers to exploring engaging public sector leadership, particularly in the Australian context (Albrecht, 2005). Thus, another

23

contribution to be addressed in this research concerns the exploration of engaging leadership in public sector organisations. Given that leadership is central to public sector efficacy, and given the need to more fully explore the role of engagement in that context, the purpose of this research is focused on establishing the specific behaviours that contribute to engaging leadership behaviours, in a public sector context. As stated earlier, the public sector is in direct competition with the private sector for talent. Further, the reforms of the late 20th century mean that many of the management behaviours have converged. The key point of difference is that the public service has limited financial means to extract higher levels of engagement. As such, non-cash and benefits drivers of engagement have greater resonance. Ultimately however, engagement, its antecedents and its consequential effects can be generalised to a variety of organisational settings and empirical research in the area will support its worth as an area to improve organisational effectiveness irrespective of ownership structure. A key objective of this study is to build on this developing research base. The role of Engaging Leadership The role of front line managers in both the public and private sectors as pivotal to employee engagement is documented in both the practitioner and research literatures (Alimo Metcalfe & Alban-Metcalfe, 2002, 2006; Frank et al., 2004; Shamir, 1995). Employees will stay if they have a good relationship and open communication with their immediate manager (Human Resources Institute, 2004). However, there is evidence that many of our leaders are not successful in this endeavour. A study of workers (Hudson Research, 2004) showed that a third of staff rated their supervisors as fair or poor. Further, Towers Perrin concluded from its

24

study of more than 85,000 people working for large and midsize companies in 16 countries, that “while many people are keen to contribute more at work, the behaviour of their managers and the culture of their organisations is actively discouraging them from doing so” (Towers Perrin, 2004, p.8). Essentially, practitioner evidence suggests that the role of the immediate supervisor or leader is of critical importance in building engagement (Towers Perrin, 2004). Employees need leaders that care about them and will help them achieve their goals and much of that engagement must be done by first line managers (Bates, 2004). An explicit consideration of relevance to this proposal is to examine the role of leader behaviours in shaping engaged employees. This thesis specifically aims to identify which domains of leadership impact on specific outcomes of engagement. In workplace research conducted by Gallup (Harter, Schmidt & Heyes, 2002), both qualitative and quantitative data have pointed to the importance of the supervisor or the manager and his or her influence over the engagement level of employees and their satisfaction with their organisations. Specifically, items measuring aspects of the organisational environment that the supervisor can directly influence explained most of the variance in job satisfaction surveys and employee opinion surveys (Harter, Schmidt & Heyes, 2002). These research findings have been replicated in individual-level meta-analyses (Judge & Piccolo, 2004) which found that the specific aspect of satisfaction most highly related to performance was satisfaction with one’s supervisor. This current research is specifically interested in the role of affect on engagement. The role of affect (the experience of feeling or emotions) in the workplace is relatively well established in the academic literature. The starting point

25

for modern research on emotion in organisations seems to have been Hochschild’s (1983) seminal work on emotional labour: The Managed Heart, and has recently experienced a resurgence of interest in the academic area (Ashkanasy, 2003; Barsade, Brief & Spataro, 2003; Brief & Weiss, 2003). However, despite the renewed interest, little empirical research exists that specifically examines affective leader behaviours and their impacts on employee engagement. What is lacking currently is a specific focus on leader behaviours that act to engender engaged responses in employees. Thus, this area of leadership, specifically affect driven leader behaviours will be examined for their impact on engagement in subordinate staff. Purpose of the Research This thesis presents a detailed examination of the construct of engagement, and specifically, an examination of affective leader behaviours in the public sector and their impact on subordinates’ levels of engagement. In forming an argument for such research, this thesis will examine literature on engagement to determine if we are to accept engagement as a new and unique concept, and in doing so, whether or not it is distinct from other more established constructs, such as organisational commitment (Meyer & Allen, 1997). A definition of engagement will be developed, and an exploration of the antecedents and consequences of engagement will be established. In particular, this thesis will provide an examination of leader behaviours as a key antecedent of engagement. Specifically, it will examine the effects of the affective (or social – emotional support) elements of leader behaviours and their impacts on the core dimensions of engagement.

26

A key aim of this study will be to develop and test a model defining the various interrelationships between affective leader behaviours, and the various dimensions of engagement. The study will draw upon research in the areas of cognitive and affective psychology to provide a directional model of engagement. The research will use Confirmatory Factor Analyses and Structural Equation Modeling to assess levels of engaging leader behaviours and their impacts on subordinates levels of engagement. Contribution to scholarship In this thesis, contribution to scholarship will be demonstrated over the following four areas. First, this thesis will clarify the competing and inconsistent interpretations of the meaning of the construct of engagement, and in providing suitable and agreed boundaries around the construct, such that it can be quantitatively and qualitatively measured. Thus one of the specific objectives of this research is to examine the literature pertaining to engagement in order to develop a definition of engagement suitable to research. Second, this thesis will determine the contributing dimensions of employee engagement with the areas of cognition, affect and behaviour. This is provided in an attempt to provide a platform for future measurement development and to provide a framework for the future measurement of the engagement construct. To that end, this thesis will develop a model of engagement, such that it can form the basis for further research in terms of measurement establishment.

27

Third, and most importantly, this thesis aims to examine the role of engaging leader behaviours in shaping subordinates’ engaged behaviours. More specifically, this research will examine the impact of affective leader behaviours and how these impact on the identified dimensions of engagement. Ultimately this research will provide a model of engagement, linking leader behaviours to attitudes and outcomes of employee engagement. This level of focus has not previously been addressed in any literature to date. While the notion of engagement has been seen to be related to all manner of constructs (e.g. job satisfaction and organisational commitment), an examination of the various relationships between engaging leader behaviour, and of the various engagement domains, has not been previously established. Finally, this research project will examine how employee engagement in the public sector is impacted by engaging leadership behaviours. Most of the existing engaging leadership research in the public sector (of which there is very little), focuses on UK and USA data, that may not be applicable to the specific circumstances of Australian organisations. Thus it is hoped that this thesis can contribute to the body of knowledge surrounding engaging public sector leadership in an Australian context. Overview of thesis structure This thesis is presented over four chapters. Chapter I has introduced the thesis. Chapter II provides an analysis of the literature relating to employee engagement and related constructs, as well as literature pertaining to affective leader behaviours. It will provide a model for testing, as well as the research questions for the thesis. Chapter III explains the methodology for data collection and analysis. The participating organisations are described, along with the sample sizes and

28

instruments used. Chapter III also presents the results for the study. Specifically, the measurement model is established in terms of the validity and reliability of the dimensions of engagement (cognitive, affective and behavioural), as well as affective leader behaviours via confirmatory factor analysis. An analysis of the structural model as well as a proposed mediation model is conducted, using structural equation modelling. Finally, Chapter IV presents a discussion of the findings, conclusions and implications of the research. Implications of the research for the public sector will also be provided. The thesis concludes with a presentation of the limitations and recommendations for further research. Conclusion This chapter has introduced and presented the rationale for the research. It has discussed the environmental and organisational contexts for engagement, illustrating the need for the construct of engagement to be clarified and refined. Specifically, this chapter has shown that the construct of engagement needs firstly to be defined, conceptualised and separated from other related and overlapping constructs (such as those of commitment and job satisfaction). Further, information pertaining to the antecedents and consequences need to be determined and rigorously tested. Of interest to this research is the role of affective leader behaviours as an antecedent to engagement. To this end, this chapter has also outlined the potential impact of leader behaviours in this regard and justified the examination of the relationship between leader behaviours and engagement. Finally, this chapter has presented the public service as a context for this research.

29

The following chapter will provide an overview of the literature in both the practitioner and applied fields of engagement and other related constructs, and will put forward the model, questions and hypotheses of interest to the research.

30

Chapter II: Literature Review The concept of a strong relationship between an engaged workforce and the financial and psychological efficacy of an organisation (Bates, 2004; Frank et al, 2004; TowersPerrin-ISR, 2006) is reason enough to conclude that the construct of employee engagement is worthy of continued investigation. While the construct itself is a relatively new one, it is heavily marketed by the practitioner community as the panacea of organisational efficacy. Academic researchers are now slowly providing informed input on the issue. To date, despite a surge in interest in improving engagement, practitioners and researchers still disagree about what employee engagement is, how to go about improving it, and what it looks like when it is achieved. The challenge for researchers is to define the construct in a way that can offer a solid basis on which to build research. To understand engagement: its consequences; its components; and its antecedents; is necessary for the efficacy of the construct. If there is no consistency in terms of its parameters, and specifically its measurable properties, both the applied and scientific fields will struggle to appropriately research and refine the construct of engagement. The following literature review will examine the construct of engagement. It will provide an analysis of the practitioner literature, with a view to establishing how it is viewed and applied by practitioners and consulting groups. It will go on to define the term and then pursue the question of whether it is a new and independent construct, or simply an amalgam of existing constructs such as: organisational commitment; job involvement; and job satisfaction. Because of the emergent nature of the construct, this review will examine engagement as an independent entity,

31

rather than as part of an established theoretical framework such as self determination theory, social exchange theory or organisational commitment. It will examine some of the potentially contributing and overlapping constructs in an attempt to define the boundaries of engagement, and to provide a platform for how it may be operationalised and measured in a valid and reliable way. This review will examine the antecedents of engagement, with specific focus on the role of leadership, and in particular; direct, or immediate leaders/supervisors. The notion of affect driven leadership behaviours in particular, will be discussed as an important antecedent to employee engagement. This review will ultimately put forward the research questions of interest in this project and hypothesize what forms of leadership are required to create engagement in subordinate staff. Why engagement matters As indicated above, the notion of engagement has acquired momentum in the applied world. As organisations are increasingly looking to human capital to provide competitive advantage over rivals and improve internal financial indicators, employee engagement is seen as one means of facilitating the utilization of human capital (TowersPerrin-ISR, 2006). To that end, engagement is heavily marketed by the practitioner community, particularly the large consulting firms (e.g., Corporate Leadership Council, 2004; Hewitt Associates LLC, 2001; Towers Perrin, 2004), as the panacea of organisational efficacy. Towers Perrin (2005) conducted a large scale global workforce study which revealed that two thirds of organisations surveyed, were suffering “significant to severe” cost pressures. This, in conjunction with ongoing corporate downsizing, the rise of off shoring and leaner organisational structures, has had two major impacts on

32

human resources (Towers Perrin, 2005). First, it has placed more strain on employees, who are being required to do more with less, resulting in unprecedented levels of work stress and dissatisfaction with employers. This in turn has resulted in an increased need for organisations to manage retention and commitment factors (Corporate Leadership Council, 2004). Second, cost pressures are intensifying the focus on human capital as a means to increase output. Given that on average, approximately 35-40% of corporate revenues are used on human capital –specifically salaries, benefits, hiring costs and training initiatives - it is logical that organisations are increasingly looking to that area of expenditure to ensure it is being used to its full capacity (Bates, 2004). Exactly how a company manages human capital expenditure is conceivably the only remaining source of enduring competitive advantage left to organisations, as other kinds of capital are readily accessible and easily copied (Nalbantian, 2006). This has put the microscope on Human Resources with the increasing expectation that it must demonstrate a return on investment (Jesuthasan, 2003). The reason engagement has hit such a chord in the business world, is due to the claims of a link between engagement and bottom line results (e.g., Frank, Finnegan, & Taylor, 2004). Engagement is marketed by the large consultancy groups (e.g., Corporate Leadership Council, 2004; Hewitt Associates LLC, 2001; Towers Perrin, 2004) as having a substantial impact both on retaining employees within a company (thereby reducing costs associated with recruitment, selection and knowledge retention), and improving employee productivity through discretionary effort, less errors, and improved customer service. Hewitt Associates LLC (2005) state that they “have established a conclusive, compelling relationship between

33

engagement and profitability through higher productivity, sales, customer satisfaction, and employee retention” (p.1). In his article “Getting Engaged”, Bates (2003) relates a story about a mortgage lending organisation that employed consultants to develop and implement strategies for engaging its staff. After engagement plans were implemented, the company tracked the revenue brought in by loan officers and found significant differences in staff productivity. Officers who rated themselves as being “actively disengaged” brought in 28% less revenue than those who were “fully engaged”. Of interest however, was that the large middle group of employees, “that show up and go through the motions but are not really engaged, and lack that extra spark” (p. 48), brought in 23% less revenue than the engaged group. Similarly, a study by ISR (2003), using data from over 360,000 employees from 41 companies, found that highly engaged companies gained a 3.81 % net profit margin (averaged over a three year period) over similar companies in which employees showed low engagement overall. A recent benchmarking report by Best Practices, LLC (2008) discussed the financial implications of employee turnover. According to data from the report, a reduction in turnover rates by just two percentage points could result in a savings of US $3 million annually in the average US Company. Depending on the level of employees (supervisors versus front-line employees) and the reduction rate, this figure could escalate to savings as high as US $40 million annually. A national salary survey in Australia (AIM, 2006) indicated that the increase in turnover was also having significant impact on Australian employers despite substantial salary increases in the previous year. The survey concluded that skill shortages had forced

34

the average national salary to increase by 4.4 per cent. The survey also found an increasing employee turnover rate was hurting many businesses. Voluntary turnover rose to 11.5 per cent, a 0.4 per cent rise on the previous year. In addition, the survey found that 71 per cent of organisations estimated that it cost them upwards of $20,000 to replace an employee, adding up to significant costs for Australian business. Gallup, a global research based consultancy group developed a research instrument in the 1990s, the Gallup Q12 Employee Engagement Index, which is administered semi annually to large numbers of employees. Meta analyses of research conducted in 7,939 business units in 36 companies, examined the relationship between employee satisfaction, engagement and customer satisfaction, productivity, profit, employee turnover, and accidents (at the business unit level). The results indicated that the reported relationships were large enough to have “substantial practical value“ (Harter, Schmidt, & Hayes, 2002). In addition, results from the 2006 Q12 review (Harter, Schmidt, Killham & Asplund, 2007) indicated that only 29 percent of employees surveyed, described themselves as being “actively engaged” in their jobs. Conversely, fifty-four percent of employees described themselves as “not engaged”. These employees had effectively tuned out, putting in the time, but without the cognitive vigilance or energy and commitment one would otherwise expect. Seventeen percent of employees described themselves as “actively disengaged”. These employees carried out their disengagement by undermining what their engaged co-workers were trying to accomplish, and their organisation’s reputations. Gallup estimated that actively disengaged workers cost US business approximately $300 billion a year in low productivity. This figure was based on two different calculations which yielded similar outcomes: (1) the total salary loss (of

35

disengaged employees) based on the $30,000 per year US average salary; and (2), the anticipated productivity gain based on an estimate of 3.41% increase in output for each of the currently disengaged employees (Harter et al., 2007). The Corporate Leadership Council (2004, 2004a) produced similar findings in their research. Based on a survey of more than 50,000 employees within 59 global organisations, CLC researchers identified the highest employee engagement impact drivers. They used these drivers to categorize employees as ‘true believers’, who demonstrate very strong commitment to their organisations; ‘disaffected’ employees, who were actively opposed to someone or something in their organisation; and the middle group of employees ‘agnostics’, who were modestly committed. CLC aggregated the results further, to determine differences between ‘engaging organisations’ (comprised of 24% ‘true believers’ and 5% ‘disengaged’) and ‘disengaging organisations’ (3% ‘true believers’ and 17% ‘disengaged’). From this data it was revealed that ‘engaging organisations’ employ 15.8% of employees demonstrating ‘discretionary effort’ and 42.9% of employees who intend to stay with the company. Conversely, the disengaged organisations had just 3% of employees showing ‘discretionary effort’ and only 15.3% of employees who intended to remain with their organisation. As argued by CLC, even moderate increases in engagement levels have a substantial impact on an organisation’s bottom line. The CLC research also highlighted other interesting aspects of engagement. For instance they concluded that the substantial differences observed between organisations indicated that engagement was not likely to be attributable to individual factors such as personality dispositions and demographic factors, but rather, indicated differences about the organisations themselves and how they

36

employ their human capital strategies. In their research, they also distinguished between rational and emotional engagement, and in this context, they found that emotional engagement was four times more valuable (than rational engagement) in terms of driving an employee’s willingness to commit to an organisation (commitment defined in terms of discretionary effort and intent to remain) (Corporate Leadership Council, 2004a). A Towers Perrin (2003) global workforce survey involving about 35,000 people working full-time for large and midsized firms across North America reached similar conclusions. In their results, only 14 percent of all employees surveyed, were ‘highly engaged’ in their jobs. The study concluded that ‘there are clear links between our respondents’ level of engagement, their focus on customers, and aspects of their organisation’s financial and operational performance across a number of areas (p. 3). Their results overall indicated that on average, 17% of the sample were ‘engaged’, 19% were ‘disengaged’, leaving a remaining ‘massive middle’ (p. 5) of 64% of ‘neutral’ employees. The latter had relatively lower agreement scores and more disagreement scores across the board. While quite positive in some areas, they had neutral to negative views on a variety of the engagement elements, particularly the emotional ones. As stated previously, this pattern was replicated across a range of demographic segments in the study. Of interest to the researchers was the fact that the traditional monetary rewards did not appear to be a significant driver of engagement. While the research indicated that factors such as salary did play a significant role in attracting people to an organisation, they had a relatively minor role at best, in driving engagement itself (Towers Perrin, 2003, p. 8).

37

A final study by Oakley (2004) found a direct link between engaged employees and customer satisfaction, and between customer satisfaction and improved financial performance. In organisations with engaged employees, it was found that customers used their products more, and increased customer use led to higher levels of customer satisfaction. In addition, customers who were more satisfied with an organisation’s products were less expensive to serve, used the product more, and hence were more profitable customers (Oakley, 2004). In summary, most of the available information (from non academic sources) indicates a strong and compelling relationship between engagement and organisational financial performance. Specifically, practitioner research has indicated a strong relationship between engagement and productivity, employee attraction and retention, and on levels of customer engagement and satisfaction. It is plausible to conclude from these studies that employee satisfaction and engagement are related to meaningful business outcomes at a magnitude that is financially significant and that these correlations generalize across companies (Harter, Schmidt, & Hayes, 2002). Furthermore, they also provide a compelling case that even relatively small increases in engagement lead to substantial direct operational gains. Notwithstanding the above, the majority of research presented thus far, has been obtained from applied practitioners and consultancy groups, rather than from academic sources with transparency applied to measures and methods of assessment. Metrics used by consultancy firms are generally not available for scrutiny, statistical methods applied to analyse the data are similarly not made available for consideration, replication or review. Thus significant questions remain unanswered with respect to construct validity, and common method variance (Harter & Schmidt,

38

2008; Podsakoff, MacKenzie, Lee & Podsakoff, 2003). Other researchers have noted similar inconsistencies, i.e. “measures of engagement we have seen in use in the world of practice are highly similar to the measures used for assessments of job satisfaction (or climate or culture), albeit with a new label” (Macey & Schneider, 2008). For these reasons, engagement as a construct must be evaluated from an academic perspective. In the following section, the construct of engagement will be scoped and defined. After which, acknowledged antecedents of engagement will be reviewed. In particular, the role of affective leader behaviours will be considered as a highly influential engagement antecedent. What is engagement? Despite its intuitive appeal, the definition of engagement is relatively unclear (Dalal, Brummel, Wee, & Thomas, 2008; Saks, 2008). While the construct has established itself firmly in the practitioner community, it remains a concept with a “sparse and diverse theoretical and empirically demonstrated nomological net” (Macey & Schneider, 2008 p.10), or one in which the exact components of engagement, its potential antecedents and consequences need to be rigorously and empirically established. This is neither unusual nor problematic in essence as many other psychological or behavioural constructs have suffered from a similar lack of precision early in their evolvement (Kanungo, 1982; Maslach & Leiter, 2008; Schaufeli, Taris, & van Rhenen, 2008), and it does not necessarily mean that the construct lacks practical or conceptual value (Macey & Schneider, 2008). Presently, however, the composition of “measured levels of engagement” that appear in various practitioner - led research is diverse and inconsistent (Harter & Schmidt, 2008). Given that these measurements are provided for the purpose of actionable outcomes,

39

it is imperative that a common understanding is reached as to precisely what is being measured, and thus assist organisations attain consistent efforts in this regard (Harter & Schmidt, 2008; Saks, 2006). The previous point will act as a guide in the explication of the construct for the purposes of this research. Ultimately it will be organisations that will benefit from a clearer picture of employee engagement. To this end, engagement as a construct must be applicable to the practitioner community, and the needs of that community must be served in the application of the science. In his book, Getting Engaged: The New Workplace Loyalty, Rutledge (2005) defines engaged employees as being “inspired by, attracted to, committed to and fascinated with their work” (p.269). Engaged employees care about the future of the company and are willing to invest the discretionary effort to ensure that the organisation succeeds (Rutledge, 2006). And they are more likely to stay with their organisations (Harter, Schmidt, & Hayes, 2002). Other definitions include: “an employee’s attitudinal attachment to his or her job and company, intention to act in company’s best interest, and willingness to invest discretionary effort in achieving business goals” (Jesuthasan, 2003, p. 57); “a person who is fully involved in, and enthusiastic about, his or her work” (Seijts & Crim, 2006, p.32); and “an individual’s involvement and satisfaction with as well as enthusiasm for work” (Harter & Schmidt, 2008 p. 269). A more individually referenced explanation from Kahn (1990), describes engagement as “the simultaneous employment and expression of a person's preferred self in task behaviours that promote connections to work and to others, personal presence (physical, cognitive, and emotional), and active, full role performances”

40

(p.694). Descriptions of positive affectivity such as: adaptive, energy, and enthusiasm (Macey & Schneider, 2008), focus and absorption (Rothbard, 2003), cognitive vigilance (Kahn, 1990), vigour, energy and dedication (Shirom, 2003), are also consistently used to define the construct of engagement, and to differentiate it from other similar constructs such as job satisfaction and organisational commitment (Macey & Schneider, 2008). Along this vein, some researchers define engagement as the opposite or positive antithesis of burnout. According to Maslach et al. (2001), engagement is characterized by energy, involvement, and efficacy, the direct opposite of the three burnout dimensions of exhaustion, cynicism, and inefficacy. Research on burnout and engagement has found that the core dimensions of burnout (exhaustion and cynicism) and engagement (vigour and dedication) are opposites of each other (Gonzalez-Roma et al., 2006). A new construct or a new blend of old wines? A recent focal article by Macy and Schneider (2008) provided an exploration of the construct of engagement, reviewing the academic and practitioner literatures and providing a conceptual model. Their discussion of engagement comprehensively reviews the engagement literature and strives to articulate the place of engagement in the context of more long standing research traditions and models (i.e. job satisfaction, organisational commitment, job involvement, positive affect and affectivity; proactive and citizenship behaviour). Much effort is spent in their treatise exploring these established constructs in terms of their partial conceptual overlap with engagement. Based on a search of the literature, Macey and Schneider (2008) provide the most comprehensive investigation and articulation of the conceptual

41

space of engagement to date, and thus will serve as the basis for the discussion of employee engagement. Macey and Schneider’s (2008) model of engagement will be evaluated in the following paragraphs with a view to determining the parameters of engagement for use in this research. Macey and Schneider (2008) offer a model of engagement that they define as an “aggregate multi dimensional construct” (p.18), as well as multi dimensional in nature. In other words, it is a re-packaging of other constructs, none of which individually, can accommodate the complexity of engagement. For example, they describe behavioural engagement as “simultaneously citizenship behaviour (OCB), role expansion, proactive behaviour, and demonstrating personal initiative, all strategically focused in service of organisational objectives” (p.19). It is this latter aspect of ‘focus’ that defines their conceptualisation. The definition that Macey and Schneider (2008) tend to adopt – is that engagement is related to vigour, dedication and absorbtion (Maslach, Schaufeli & Leiter, 2001). Engagement is portrayed overall as an active performance concept, and that the different constructs that overlap with engagement, are different: in terms of their referent (job, individual or organisational); and in terms of their focus or outcome (organisational efficacy). The Macey and Schneider conceptualisation of engagement has researchers and practitioners having organisational competitive advantage as the starting point in thinking about engagement, which is a viewpoint with considerable merit.

42

F Figure 1: Macey M and Schneider, S (2 (2008) modeel of employyee engagem ment

Specificallyy, the modell (see figuree 1), is show wn as comprrising three different concepttualisations:: trait engaggement (or personal p disspositional elements); e psychollogical “statte” engagem ment, encom mpassing eleements suchh as affectivve satisfacttion, comm mitment and involvemennt); and beh havioural enngagement – notably extra-roole behaviouur and elem ments of orgaanisational citizenship c behaviour ((OCB). Macey and a Schneidder (2008) also a indicateed that these conceptuaalisations shhould be separateed from the antecedents of engageement (to bee dealt with later) and thhe consequuences of enngagement - which are of value to organisatioons. The moodel also shows thhat conditioons of workk (aspects off the job chaaracteristicss model) as well as leadershhip, also havve both a diirect and inddirect effectt on state annd behaviouural engagem ment. Transsformationaal leadershipp in particullar is repressented in thee model as having a direct effeect on trust and an indirrect effect through t the creation off trust on behaviooural engageement (Maccey & Schneeider, 2008)). This latteer aspect is drawn from Kaahn’s (19900) conceptuaalisation of engagemen nt, and was an a aspect off much

43

earlier works on leadership and trust (e.g., McGregor, 1960). The role of leadership (particularly transformational leadership) on engagement will be taken up as a separate point further in this chapter. Overall, the model has provided an excellent attempt to distinguish the construct space of employee engagement (Delal et al., 2008), and has made a significant contribution to identifying the facets of existing scholarly constructs (involvement, satisfaction and commitment) that overlap with the engagement construct. It has made advancements towards its objective of bridging science and practice (Macey & Schneider, 2008). However, in some respects, the model as proposed by Macey and Schneider is inconsistent with other interpretations of engagement, resulting in some areas of confusion. Specifically, areas of concern within the model include: their conceptualisation of state engagement; the contribution of trait engagement, and the composition of behavioural engagement. Each of these points will be addressed below. A criticism of psychological state engagement as proposed by Macey and Schneider (2008) is directed at their decision to combine the cognitive and affective aspects of engagement into a common dimension of state engagement. This is contrary to other stipulations of engagement. For instance, in Kahn’s (1990) view, engagement is a three-dimensional construct in which people can be emotionally (affectively), cognitively and physically engaged. In this view, cognitive engagement refers to those who are rationally aware of their role and mission within an organisation; emotional (affective) engagement refers to those who are emotionally connected to others (such as managers and co-workers); and physical

44

engagement refers to physical expression of engagement, such as extra-role performance (Kahn, 1990). In his theory, Kahn (1990) stipulates that it is possible to be engaged on one dimension (cognitive or affective) and not the other, though he would say that a synergistic affect would apply in that the higher the levels of engagement on each dimension, the higher overall the level of personal engagement. Despite the lack of clarity between the two (affective being arguably the emotional manifestation of cognitions), it is relevant to separate out these aspects of engagement, because a point of interest to researchers is in distinguishing the scope of the various dimensions of engagement, as well determining the existence of a directional sequence. For instance, it has been postulated that cognitive engagement is a precursor to affective engagement, and it may also be that different antecedents act as different triggers for the two (Harter & Schmidt, 2008; Towers Perrin, 2003). For this reason, separate constructs of cognitive / affective engagement will be used for this research, along with behavioural engagement. This distinction will be used in much the same way that Kahn’s (1990) conceptualization of engagement has been articulated. Macey and Schneider (2008) also proposed that trait engagement (a precursor to state engagement) consists of several personality based constructs including: autotelic personality (purpose for existence within the self); trait positive affectivity; proactive personality; and conscientiousness. These behaviours all suggest a tendency to “experience work in positive active and energic ways and to behave adaptively” (p. 21). This proposition is provided on the basis of previous research (Schneider, 1995) that suggests that people who have positive dispositions, and who

45

demonstrate engaged behaviours are more likely to choose to work in environments where they are supported to do so. In Macey and Schneider’s (2008) view, both state engagement and behavioural engagement are at least partially the result of dispositional influences. When viewed in this way, essentially as an antecedent to state engagement, the explanation becomes troublesome. The antecedents of engagement are numerous and extend well beyond dispositional attributes and include: the conditions under which people work (Kahn, 1990; Maslach & Leiter, 2008), leadership attributes (Alimo-Metcalfe & Alban-Metcalfe, 2002); as well as organisational culture (Harter et al., 2002), among others. Clearly, the antecedents of engagement are of great interest to organisations and those who wish to leverage the construct of engagement. For this reason, dispositional attributes may be considered but one of a group of antecedents, including job and organisational characteristics, leadership skills and dispositional influences. Thus, for the purposes of this research, trait engagement will not be considered as a distinct dimension of engagement, but rather an antecedent to engagement. In summary, for the purposes of this research, engagement will be conceptualised as a three-dimensional construct in which people can be cognitively, affectively and behaviourally engaged. This conceptualisation specifically takes the notions of state and behavioural engagement from the Macey and Schneider (2008) model. In addition, it separates out the cognitive and affective components of “state” engagement, consistent with Kahns (1990) conceptualisation of engagement and treats them as two distinct dimensions. The following section will deal with each of these dimensions in terms of their role in the overall model of engagement proposed for this research.

46

F Figure 2: Three Th dimenssional construct of eng gagement

Cognitive Engagement E t T term “ccognitive” refers The r to an information n processingg view of ann individuual's psychoological funnctions (APA A, 2006). Research R coonducted byy employee researchh consultanccy firm, ISR R (2008), reevealed that the "cognittive" engageement of employeees - whethher they suppport and unnderstand thee company's strategy aand the direction it is takinng - was as important i too organisatio ons as was affective a enngagement and specifically, thhat it was ann equally strrong a driveer of financiial performaance. C Cognitive e engagement t is not a weell defined construct c in either the aapplied or academ mic literaturees. Of all thee proposed dimensionss of engagem ment, it is thhe least well deffined and sccoped and consequently c y has no asssociated meeasure. Kahhn (1990) expressees it (indireectly) as a nootion of liveely awareneess, intellecttual vigilance and a sense off heightenedd perceptionn and intereest. In much h of the pracctitioner liteerature, it is referrred to as “raational engaagement”, orr the intelleectual comm mitment peoople have to their organisations (Corporaate Leadershhip Councill, 2004; Tow wers Perrin,, 2003). t notion of o ‘intellectuual commitm ment’, cognnitive engagement In breakking down the might also be desccribed as an understandding of goal and role clarity (Pricew waterhouse Coopers C 20002; Sawyer,, 1992); an understandi u ing of the

47

organisation’s objectives, successes and failures – referred to as Line of Sight (Hillman, 2007; Towers Perrin, 2003); as well as an outcome of a rational assessment of organisational information in terms of the implications for the individuals wellbeing (Corporate Leadership Council, 2004; Lazarus et al., 1980; Towers Perrin, 2003). These notions will be discussed respectively. Cognitive (rational) engagement would appear to be largely an individual’s understanding of their role in the context of the organisation’s goals and objectives. As argued by Steers and Porter (1991), if employees see themselves as effective contributors to the company's goals or objectives, they are more likely to perform at a higher level. Job meaning in this context can be articulated through Hackman and Oldham’s (1976, 1980) job characteristics model. The model proposes that a job’s skill variety, task identity and task significance combine to account for the meaningfulness of a job. Cognitive engagement, defined in terms of understanding how to contribute to an organisation’s objectives, is arguably consistent with employees understanding of why their contribution matters or how it affects others i.e. task significance (Boswell 2006). Employees who understand how to contribute to an organisation’s strategic goals are more likely to feel a sense of belonging (or fit). Prior research in this area: specifically, job design (Hackman & Oldham, 1975); role clarity (Jackson & Schuler, 1985), and perceived fit (Cable & Judge, 1996) indicate that employees want to see how they fit with or contribute to the organisation. This suggests that cognitive engagement may be instrumental to employee performance and retention. Thus there are multiple reasons to expect that aligning employees with the strategic objectives of the firm is an important concept for HR professionals and academics to investigate.

48

The concept of Line of sight (LOS) is related to goal and role clarity, but focuses more on the processes by which goal and role clarity are achieved. In particular, it refers to how organisations can communicate their goals and objectives in a way that contextualizes individual endeavours. It enables an employee’s understanding of the organisation’s goals and the actions that are necessary to contribute to those objectives (Boswell & Boudreau, 2001). The theory behind LOS conjectures that employees are more likely to feel a sense of personal ownership for the business results if they know what the organisation’s objectives are, and how the work they do on a daily basis actually contributes to those objectives and strategies. LOS has been postulated to significantly impact upon organisational efficacy (e.g., Hatch & Dyer, 2004). Employees who are aligned with the firm’s goals become engaged in tasks and behaviours that foster tacit learning; essentially, learning by doing. “When human resources are aligned with and possess tacit, firm-specific knowledge, employee behaviours associated with the firm’s specific strategic goals become a source of competitive advantage: valuable, rare, and leading to increased firm performance” (Boswell 2001, p. 501). In its global human capital survey, PricewaterhouseCoopers (2002) suggested that successful organisations have recognised the importance of effectively delivering business strategies and the policies and practices that accompany them, directly to individual employees. This is reiterated by Towers Perrin (2003) who claim that specifically communicating the connection between the individual and the broader organisation is “paying off and helping employees more clearly understand the mutual responsibility and accountability at the heart of the employer/employee relationship” (Towers Perrin, 2003 p. 11) . Similarly, the Corporate Leadership Council (2004) found that ‘Line of sight’ or a clear connection between an

49

employee’s job and organisational strategy was a vital component of employee engagement. According to Boswell (2006), LOS is not merely about whether employees perceive they are contributing effectively. Research conducted by Towers Perrin (2003), found that many organisations confuse communication with information, concentrating on disseminating basic facts rather than providing context, commentary and two-way dialogue. Their research indicated that employees want to know what management thinks and believes and how it plans to act. And they also want the opportunity to provide input - “it is part of the environment of mutual trust, accountability and responsibility that is important in engaging people and winning discretionary effort” (Towers Perrin, 2003, p. 11). LOS provides transparent links between individual capability and business growth strategies. The more transparent managers can make the firm's operations, the more effectively employees can contribute to the firm's success. Transparency is important because it helps employees see the link between their actions and the performance of the firm, thereby enhancing the cognitive aspect of engagement (Konrad, 2006). Increased understanding of organisational goals should lead to improved outcomes, as employees will be more likely to do the right thing (Kristof, 1996). Similarly, a lack of understanding may adversely affect performance as workers may inadvertently work on low-priority goals (Witt, 1998). Finally, cognitive engagement would appear to cover not only the understanding of goal and role clarity and LOS, but also an outcome of a rational assessment of organisational information in terms of the implications for individual well-being (Corporate Leadership Council, 2004; Lazarus et al, 1980; Towers Perrin,

50

2003). Rational appraisal of ‘personal benefit’ may extend beyond a sense of goal and role clarity to a belief that the organisation is likely to provide opportunities for employees to contribute to organisational goals, thus creating the possibility of empowerment, contribution and potential advancement (Corporate Leadership Council, 2004). Thus, cognitive engagement for the purposes of this research is goal and role clarity, contextualised by a clear line of sight to the organisation’s overall goals and objectives, thus providing a clear sense of job meaning. In addition, it refers to a rational appraisal by individuals that the organisation is likely to meet his/her objectives in terms of contribution, support and advancement. The following section will deal with affective engagement, hypothesised to be a consequence of cognitive engagement, and an antecedent of behavioural engagement. Affective Engagement The previous section dealt with the role that cognitive engagement plays in overall employee engagement. This section explores the second important component of the engagement construct, that of affective engagement. Affective engagement refers to the experience of feeling or emotional connection to one’s organisation; or something or someone within it (Kahn, 1990). The term "affective" infers an instinctual reaction to stimulation. Many theorists (e.g., Lazarus, 1982) consider affect to be post-cognitive: In this view, an affective reaction, such as liking, disliking, trust or commitment, is based on a prior cognitive process in which a variety of content discriminations are made and features are identified, examined for their value, and weighted for their contributions (Brewin, 1989). For example, Lazarus et al. (1980) argue that an observer’s emotional

51

reactions to a target are an outcome of cognitive appraisals. Specifically, ‘‘What are the implications of this information for my well-being”? Affective engagement can be seen as a broad multi-dimensional construct embracing a family of related and more specific constructs focused on individuals’ relationships with their work roles (Maslach & Leiter, 2008; Salanova, Agut, & Peiro, 2005). Macy and Schneider (2008) refer to it as “a new blend of old wines with distinct characteristics and feel” (p.10). In particular, it overlaps conceptually (and significantly) with organisational commitment (Corporate Leadership Council, 2004; Meyer & Allen, 1997; Porter, Steers, Mowday, & Boulian, 1974; Wellins & Concelman, 2005); job involvement (Lawler & Hall, 1970; Lodahl & Kejner, 1965); and job satisfaction (Burke, 2008; Harter et al., 2002; Kanungo, 1982). It can also be seen in the areas of trust and perceived organisational support (POS). For example, studies by Laschinger et al. (2000) and Tan and Tan (2000) conclusively demonstrated that increased trust in management influenced employees’ levels of affective commitment. Furthermore, and again consistent with the role of affect on organisational outcomes, Connell et al. (2003) reported that trust in management, rather than directly influencing extra role performance (ERP), indirectly influenced ERP through its influence on affective commitment. These overlapping areas will be discussed below, in terms of their contribution to engagement. Organisational commitment can be described as the relative strength of an individual's identification with, and involvement in, a particular organisation. Conceptually, this construct can be characterised by at least three factors: (a) a strong belief in and acceptance of the organisation' s goals and values; (b) a willingness to exert considerable effort on behalf of the organisation; and (c) a strong desire to

52

maintain membership in the organisation (Mowday, Porter, & Steers, 1982). Prior research supports the notion that commitment to the organisation is related to a number of desirable employee outcomes. For example, organisational commitment has been linked to extra-role behaviour (Gregersen, 1993; Shore & Wayne, 1993), job performance (Mathieu & Zajac, 1990), and lower turnover (Bishop, Scott, & Casino, 1997; Mathieu & Zajac, 1990). Macey and Schneider (2008) argue however, that organisational commitment and other pre-existing constructs in their totality cannot explain engagement, but rather aspects of each of them – specifically the affective, energetic elements contribute to the notion of psychological state engagement. This is intrinsically different they argue, from what might otherwise be another manifestation of a same latent job attitudinal construct. By way of example, the authors argue that engagement contains the affective and energic/arousal components of the job satisfaction construct, but not the contentment or satiation aspects – which they argue is the fundamental intention of the job satisfaction construct (Macey & Schneider, 2008). Shaufeli (2006) also argues that engagement is different from job satisfaction expressed as engagement, the former being primarily about positive affect and activation (i.e. feelings of energy and enthusiasm). In similar vein, Kahn (1990), also asserts that employee engagement is different from other constructs such as job involvement, commitment or intrinsic motivation because those constructs are too remote from the daily world of work, the nature of organisational efficacy and people behaving in relation to specific work environments. Thus, employee engagement may be seen as more proximal to the work experience than other constructs such as organisational commitment, which are more distal in their conceptualisation.

53

In contrast to the above mentioned constructs, it has been argued that engagement focuses on the psychological and behavioural experiences of work, and the nature of those particular work experiences act to influence the degree to which people present or absent themselves in work experiences (Kahn, 1990; Luthans & Peterson, 2002). Existing organisational behaviour concepts focusing on person-role relationships emphasise the generalised (rather than contextual) states that employees occupy. The key difference for the construct of engagement, is how it is extracted as “positive attachment to the organisation entity and measured as a willingness to exert energy in support of the organisation” (Macey & Schneider, 2008, p. 9). In summary, in much the same way that milk may be considered and applied differently to butter (though derived from the same source), so it is with engagement. Although the construct of engagement overlaps with other constructs, its composition, its style of extraction and application, justifies its place as a distinct and unique construct, consisting of cognitive and emotional components, and distinguish it from other related constructs most obviously; organisational commitment and job satisfaction (Macey & Schneider, 2008; Saks, 2008). Affective engagement, while it overlaps with other constructs, for the purposes of this research is seen as a distinct and unique construct, and as distinguishable from other related constructs such as organisational commitment, job satisfaction and job involvement. Further, it is postulated to be post cognitive, that is occurring after cognitions have taken place. As will be discussed in the following section, it is also proposed to precede behavioural engagement.

54

Behavioural engagement Macey and Schneider (2008) describe behavioural engagement, as “adaptive behaviour intended to serve an organisational purpose” (p.18). This dimension of engagement is seen as directly observable behaviour in the work context and includes innovative behaviours, initiative, proactive behaviours and going above and beyond what might otherwise be expected. In coming to this conclusion, they preface three streams of thought; that of Organisational Citizenship Behaviour (OCB) and its variants; role expansion (Crant, 2000) and personal initiative (Frese & Fay, 2001). From this viewpoint, behavioural engagement is seen as adaptive behaviour. This viewpoint is consistent with Kahn’s (1990) position on the behavioural manifestation of engagement. Kahn specifically emphasised the adaptive requirements of modern organisations, and described engaged employees as those who were able to adapt, take initiative and be responsive in ever changing circumstances. Kahn (1990) and others (e.g. Maslach & Leiter, 2008; Saks, 2008) place a caveat on the notion of discretionary effort, and argue that behavioural engagement refers to how well employees express themselves vigilantly and competently within their roles and are “psychologically present” (Kahn, 1990), rather than the notion of doing more than expected. Interestingly, Macey and Schneider (2008) did not address the notion of turnover in their discussion of behavioural engagement, despite their positioning of behavioural engagement as “strategically focused and bounded by organisational purpose and relevance” (p.18). This is contrary to most other portrayals. Most of the practitioner researchers (e.g., Corporate Leadership Council, 2004; Towers Perrin, 2003) identify a combination of behavioural commitment and discretionary effort. This is variously expressed as: pride in the organisation and willingness to extol its

55

virtues; and intent to remain with the organisation (Catteeuw, Flynn, & Vonderhorst, 2007; Harris & Cameron, 2005; Harter et al., 2002; Saks, 2006); as well as discretionary effort, that is, being willing to go above and beyond the normal role requirements (Masson et al., 2008; Towers Perrin, 2004). Academic research conducted in related organisational behaviour fields also informs the dimension of behavioural engagement. For example, social exchange theory explains how employees, if they perceive that their organisation values their contribution and cares about their well-being, develop a ‘felt obligation’ (Eisenberger et al., 2002) to ‘reciprocate’ (Gouldner, 1960) with pro-social attitudes and positive behaviour in the form of increased commitment, extra-role behaviours, and decreased turnover (Armeli et al., 1998; Eisenberger et al., 1986; 2002). Other researchers contend that employees who feel an emotional attachment to their organisation will likely demonstrate a willingness to help (Bishop & Scott, 2000), act with the best interests of their organisation at heart, and be less likely than others to exit their organisation (Meyer & Allen, 1997). In summary, behavioural engagement is a worthwhile construct and of direct benefit to organisations, thus it will be a core dimension of the engagement construct to be used by this research. Although Macy and Schneider (2008) assert that there is inconsistency in terms of how the applied world determines behavioural engagement, and thus measures it, it has been argued that there is more consistency than perhaps they acknowledge (Masson et al., 2008). Accordingly, this research will use the construct of behavioural engagement as defined by Macey and Schneider (2008) consisting of elements of organisational citizenship behaviour and extra role behaviour, but will add the component of intention to remain with an organisation, in

56

keepingg with other researcherss in the fieldd. The three proposed dimensionss of employeee engagem ment for thiss research arre outlined in i Figure 3..

F Figure 3: Dimensions D of engagem ment and keyy associatedd constructss D Definition o engagemeent. of I the preceeding sectioon, the compponents of engagement In e t were identtified, as shown in i Figure 3. The breaddth of the coonstruct need ded to be exxplored in oorder to define itt in a more grounded context. c Thiis next sectiion will estaablish a worrking definitioon of the terrm engagem ment. A Although d different authhors use diffferent term ms to describbe employeee engagem ment, there are consisteently occurrring themes as follows:: that emplooyee engagem ment is a deesirable condition; has positive p imp pacts on orgganisationall efficacy, and connnotes invollvement, com mmitment, passion, en nthusiasm, focused fo effoort, and energy. The anteceedents of suuch attitudess and behav viours are deetermined bby the conditioons under which w peoplee work (Kahhn, 1990; Maslach M & Leiter, L 2008). Overall, engagem ment is conssidered to be b of value to t organisattional effecttiveness, parrticularly

57

in terms of discretionary effort and intention of employees to remain with their organisations (May, Gilson, & Harter, 2004). Employee engagement has been defined as a three dimensional construct, with cognitive, emotional, and behavioural components (Macey & Schneider, 2008; Kahn, 1990; Saks, 2006), and will be so defined for the purposes of this research. The cognitive dimension of employee engagement refers to employees who are cognitively vigilant and contextually aware of their roles and goals within their organisation, and possess an understanding of the organisation and its overall goals and objectives. The emotional (affective) dimension refers to how emotionally connected to others-such as managers and co-workers, how employees feel about the organisation, its leaders, and working conditions and whether they have positive or negative attitudes toward the organisation and its leaders. The behavioural dimension of employee engagement consists of the discretionary effort engaged employees bring to their work in the form of extra time, energy and cognitive vigilance (Kahn, 1990), and how long they remain with their organisations. The following definition from Jesuthalen (2003) is the one that will be used to guide the research that follows. In summary it provides an individual level of referent, it is comprised of both attitudinal and behavioural engagement dimensions, and is framed in terms of organisational efficacy. “Employee engagement is an employee’s attitudinal attachment to his or her job or company, intention to act in the company’s best interest and willingness to invest discretionary effort in achieving business goals” (Jesuthalen, 2003, p. 57). Attention will now turn to the factors that create engagement. Understanding the antecedents or drivers of engagement is of great interest to organisations

58

interested in leveraging engagement to produce employees who work more productively, are retained longer, and consequently contribute to more profitable and competitively sustainable organisations. The antecedents (or drivers) of engagement. The purpose of the previous section was to articulate the components of engagement. Attention will now turn to the antecedents (drivers) of engagement. This is of interest to theorists and practitioners alike, and particularly to organisations who are increasingly focused on identifying the organisational elements which are able to effect sustainable competitive advantage. While there is a significant body of research conducted in the practitioner community on the antecedents of engagement, there is limited empirical academic research in the area. Nevertheless it is possible to identify a number of potential engagement antecedents from the academic and applied literatures, which tend to fall generally to two broad areas: affective factors, specifically, social and emotional support factors; and cognitive (rational appraisal) factors. Of particular note is the role of transformational leadership as an important antecedent to all three dimensions of engagement. The role of transformational leadership was also addressed as an important component of the Macey and Schneider (2008) model of engagement (see figure 1). Bates (2004) argues that the root of engagement is not in the cognitive (rational) realm, not about how much people are paid, but more about how they feel. He indicates that managing an employee’s emotions is central to managing engagement and that ultimately engagement motivates people’s desire to give

59

discretionary effort and to stay in their jobs. However, most commentators in the field agree that engagement is in essence, a combination of cognitive and emotional needs. Kahn (1990) concluded that employees become emotionally and cognitively engaged when they know what is expected of them, have what they need to do their work, have opportunities to feel an impact and fulfilment in their work, perceive that they are part of something significant with co-workers whom they trust, and have chances to improve and develop (Kahn, 1990). A central tenant to Kahn’s model of engagement is the notion of psychological safety, which is a consequence of the amount of care and support employees' believe are provided by their organisation as well as their direct supervisors (Kahn, 1992). Employees feel safe in work environments that are characterised by openness and supportiveness. Such environments allow employees to experiment, to try new things and risk failure, without fear of the consequences (Kahn, 1990). According to Kahn (1990, 1992), engagement can also be attained from an occupational environment that is challenging, has variety, and enables the use of different skills, personal discretion, and the opportunity to make important contributions. Kahn draws comparisons with Hackman and Oldham’s (1980) job characteristics model, specifically the five core job characteristics (i.e. skill variety, task identity, task significance, autonomy, and feedback). He concludes that jobs that are rated highly on the core job characteristics provide individuals with the opportunity to bring more of their ‘self in role’ (p. 692) into their work or to be more engaged (Kahn, 1992).

60

In their empirical test of Kahn's model, May et al. (2004) replicated Kahn’s findings and found that meaningfulness, safety, and availability were all significantly related to engagement. In particular, they found that rewarding co-worker and supportive supervisory relationships were positive predictors of psychological safety, while adherence to co-worker norms and self-consciousness were negative predictors. They also found that job enrichment and role fit were positive predictors of meaningfulness; and the availability of resources was a positive predictor of psychological availability. From a burnout perspective Maslach, Schaufeli and Leiter (2001) also determined that the antecedents of engagement were derived from a central tenant of social and emotional support as provided by the supervisor and the organisation. In fact, a lack of support from supervisors was found to be an especially important factor linked to burnout. They found that recognition and reward was important for engagement, and conversely, a lack of rewards and recognition could lead to burnout. Fairness and justice was also one of the significant work conditions in their engagement model. Where fairness was perceived to be low, burnout was exacerbated, while positive perceptions of fairness increased engagement. The workload and control conditions from the Maslach et al. (2001) engagement-burnout model also suggest the importance of job characteristics, especially feedback and autonomy, in promoting engagement. Finally, Saks (2006) tested a model of the antecedents and consequences of job and organisation engagements based on social exchange theory (SET). He found that employees who were more likely to be engaged perceived they had higher levels of organisational support; had higher scores on job characteristics; and had higher

61

perceptions of procedural justice. Engaged employees were more likely to have a high-quality relationship with their supervisor which resulted in them having more positive attitudes, intentions, and behaviours. In other words, when employees believed that their organisation (and their supervisor) was concerned about them and cared about their well-being, they were more likely to respond by attempting to fulfil their obligations to the organisation by becoming more engaged. First-line supervisors in particular, were believed to be especially important for building engagement and to be the root of employee disengagement, because employees tend to view their supervisor's orientation toward them as indicative of the organisation's support (Eisenberger et al., 2002; Rhoades & Eisenberger, 2006). In summary, from the perspective of empirical research, the antecedents of engagement for employees include some rational factors such as perceived sense of fairness as well as recognition and reward. Of significance however, appears to be the affective factors of social and emotional support - particularly perceived supervisory and organisational support, and the provision of a psychologically safe and supportive environment in which a person is enabled to bring “self into role” (Kahn 1990, p. 700). Considerable research in the practitioner community has also identified the antecedents to engagement (to be explored in the following section). As will be seen, there is considerable overlap between academic and practitioner researchers in terms of engagement antecedents. Through their research, Towers Perrin (2004) found that emotional factors such as feeling cared about and being supported, directly contributed to employees’ personal satisfaction and the sense of inspiration and affirmation they get from their work and from being part of their organisation (p. 4). They state that “retention and

62

engagement are broadly similar and emphasise the emotional side of the equation” (p. 7). In contrast, they found that the antecedents of attraction (recruitment) were mostly rational (Towers Perrin, 2005). Towers Perrin (2003) identified the following four areas as the main contributors to employee engagement: Compensation and benefits; Organisational culture (collaborative & flexible, good citizenry and good connections with other groups within the organisation); resources (skills, capabilities and equipment); and above all, leadership qualities. Corporate Leadership Council (2004) research on engagement identified engagement antecedents as also centering around rational and emotional factors, but specifically found that emotional engagement was four times as valuable as rational engagement in predicting an employee’s willingness to commit to an organisation, with one exception. They found that ‘line of sight’ as discussed earlier, was the most important antecedent of employee engagement (though this depended on the abilities of leaders) . Other points of interest in their research, showed that managers, particularly first line supervisors, played a crucial role in directly driving employee commitment, and that managers were most important as enablers, or conduits of emotional engagement to their jobs, teams and organisations (Corporate Leadership Council 2004, 2004a). In respect of the antecedents to engagement as explored by the academic and research sectors, several common themes emerge. Most particularly is the overriding importance of the affective experience of work. Employees are not apathetic or indifferent to their organisations and in fact, tend to have strong emotions about their work. From the information provided, employees relate to work on a very personal level defining the total work experience in terms of what Towers Perrin (2003) have

63

explained as three “my” categories. First is “myself”: the extent to which work gives employees a sense of confidence, competence and control over their futures. Second is “my job”: the nature of what they do, how they contribute, and how they are recognised and rewarded. And last is “my workplace”: the people they work with and for (perceived organisational and supervisory support), as well as the company’s culture and atmosphere. A final point to be made in terms of the drivers of engagement is that many of them have to do with leadership. Simply put, “it’s impossible to underestimate the role leadership plays in building an engaged workforce” (Towers Perrin, 2003, p. 8). CLC (2004) similarly contend that leaders are the primary conduits of employee engagement. The next section of this report will specifically examine the role of leadership in engagement, and its impact as a driver or antecedent of engagement. The role of engaging leadership There is a substantial body of literature dedicated to the study and practice of leadership. However, it is the convergence of leadership and engagement that is the area of interest for this research. In particular, an examination of the behaviours and characteristics of managers who achieve high levels of engagement in their staff is the subject of interest. While the notion of engaging leadership is a new area of research, many areas of empirical research and applied practitioner findings partly cover aspects of this collaboration, most notably within the transformational leadership literature, and within the role of affect in organisational behaviour, both of which will be discussed below. Suffice to say however, the role of leadership in the context of engagement appears to be a conduit of enormous proportions. The role of engaging leadership will be discussed in the following section. As will be seen, the

64

direct linkages between leadership and engagement have been discussed by the practitioner community, but to date, this is a new and emerging area for academic research, and one that is open to a broad array of empirical research possibilities. Direct leaders as the conduits of cognitive engagement Cognitive engagement would appear to be highly dependent on the communication strategies/behaviours of the sender. This highlights the critical role of immediate supervisors/leaders as the conduits or enablers of organisational meaning, as they are effectively in control of the many aspects of communication that provide the transparency between individual capability and business growth strategies. Put another way, the capacity to influence cognitive engagement exists predominantly in the capabilities and behaviours of direct supervisors. They, more than any other organisational entity have the access and resources to enable employees to understand how their job specifically fits in with the organisation, and therefore increases their sense of meaning. Research conducted by Harris (2007) showed that the effectiveness of leaders' communication had a significantly strong relationship to employee’s engagement and intent to stay. More specifically, their research found that each level of leadership and message communicated by that level revealed a correlation to employee engagement and intent to stay. Additionally, the study examined the relationship between employee engagement and intent to stay with several significant leader communication styles. More direct one-on-one communication aimed at specifically linking employee roles and behaviours to larger organisational goals was the key to LOS. Such one-on-one communication was also identified as more likely to reveal inconsistent perceptions or behaviours relative to strategic objectives and allow for clarification and reorientation (or two way dialog). This personalised approach to strategic communication suggests a particularly

65

important role for the direct supervisor. Highly interactive managers who believe in the abilities of their subordinates and provide them with information and resources are more likely to have aligned employees (Boswell 2006b). The capacity to create understanding between individuals and an organisation’s strategic objectives is rooted in their immediate leader’s communication behaviours. Specifically: ensuring information is communicated with their subordinates in a meaningful, reciprocal way; encouraging employee participation in decisions that affect the organisation and their jobs; and implementing new employee socialization efforts aimed at clearly linking employee behaviours to the organisation’s success; are considered critical engagement skills. For this reason, direct leaders are considered to be a significant driver of cognitive engagement, and will play a significant role in the pathway to behavioural engagement. How management practitioners view leadership and engagement. Management practitioners have for some time been touting the importance of leadership in the engagement process. The oft quoted ‘people join companies, but leave managers’ (Towers Perrin, 2005, p. 8) underpins this position. In determining the specific drivers or leavers of engagement, research by many of the global consultancy firms (e.g., Gallup, Towers Perrin, Corporate Leadership Council) have reinforced the role of leadership as a driver of engagement, and how it operates as a filter for many of the identified levers. For example, Towers Perrin (2003) identified the following factors (in order of priority) as the most important levers of engagement. •

Management’s interest in employees’ well-being

66



Challenging work



Decision-making authority



Evidence that the company is focused on customers



Career advancement opportunities



The company’s reputation as a good employer



A collaborative work environment where people work well in teams



Resources to get the job done



Input on decision making



A clear vision from management about future success. As can be intuitively inferred, the absence or presence of elements such as

those mentioned above depends substantially on the skills and behaviours of the direct supervisor or manager. Towers Perrin (2003) concluded from its report, that “while many people are keen to contribute more at work, the behaviour of their managers and the culture of their organisations is actively discouraging them from doing so” (p. 9). Moreover, the USA Society for Human Resource Management found in its 2005 Workplace Productivity Survey, that almost six out of ten (58 per cent) Americans identified poor management as the biggest obstacle to productivity (SHRM, 2005). In Towers Perrin (2003) research, manager behaviours identified as having the most significant impact on employee engagement (in descending order of importance) were as follows; •

Supporting teamwork



Acting with honesty and integrity



Encouraging/empowering people to take initiative



Encouraging new ideas and new ways of doing things

67



Having valuable experience/expertise



Overall quality of supervision



Providing clear goals and direction



Inspiring enthusiasm for work



Ensuring access to a variety of learning opportunities



Helping employees understand how they impact financial performance



Building teams with diverse skills and backgrounds The majority of the items listed above have been repeatedly put forward by

the practitioner community as imperative to engaging staff. For example, in workplace research conducted by Gallup (Harter, 2004), research data indicated the importance of the supervisor or the manager and his or her influence over the engagement level of employees and their satisfaction with their organisations. Specifically, they found that items measuring aspects of the organisational environment that the supervisor could directly influence, explained most of the variance in lengthier job satisfaction surveys and lengthier employee opinion surveys. These research findings have also been replicated in individual-level metaanalyses (e.g., Judge et al., 2001), who found that the specific facet of satisfaction most highly related to performance was satisfaction with one’s supervisor. In essence, the way in which the direct supervisor goes about managing, significantly influences the subsequent level of subordinates’ engagement. It is now incumbent upon management researchers to replicate empirically, the premise that consultancy firms operate upon. Empirical research to date has not explicitly dealt with engaging leadership behaviours and their role in influencing levels of employee engagement;

68

however, as can be seen in the following section, inferences can be drawn from the existing academic literature. As will be seen, affective leadership behaviours that act to manage the emotions and sense of meaning that employees give to their jobs, is at the foundation of engagement antecedents. Academic research on leadership and engagement. A large body of empirical research has linked leadership with follower thinking, feeling, and behaviour (Albrecht, 2005). Bryman (2006) for example referred to a “new paradigm” view of organisational functioning and leadership which explicitly acknowledged that generating employee commitment and motivation, rather than obligating employees to follow, is a more effective and productive way to achieve organisational success. Within the context of organisational outcomes, researchers have variously discussed traits or competencies possessed by leaders (Timmerman, 2006); situational approaches to leadership (Ashkanasy,1991); behaviours of leaders (Barbuto, Fritz, Matkin & Marx, 2007); and contingency (or interactional) approaches to understanding leadership (Yammarino, Dubinsky & Comer, 1997). Each of these streams of research has contributed to the understanding of leadership and leadership behaviours. However, despite the contribution of these various approaches, an understanding of specific leader behaviours that pertain to engagement outcomes remains vague and empirically ill established (Albrecht, 2005; Frank et al., 2004; Taylor, 2004). In particular, an understanding of the way in which leaders behave that can engender an affectively loaded or emotionally engaged response in employees is of particular interest (Taylor, 2004). Fox and Spector (1999) have called for more research in how affective experiences influence behaviour and outcomes in organisational settings.

69

Also, George (2000) has argued that the affective or emotional dimensions associated with leadership have not been sufficiently researched and documented. We will now look at the two academic streams of most relevance to the notion of engaging leadership behaviours, namely the role of “affect” at work, and transformational leadership. The role of Affect in leadership and engagement The role of affect (the experience of feeling or emotions) in the workplace is well established in the academic literature. The starting point for modern research on emotion in organisations seems to have been Hochschild’s (1983) seminal work on emotional labour: The Managed Heart. This work inspired Rafaeli and Sutton's work (1987, 1989) which directed the attention of academic researchers to emotional expression as part of the work experience. Research in the area of affective events theory (AET), provided information that indicated that the even apparently insignificant irritations and positive events that people experience each day at work, accumulate to determine organisational members' affective states, and these states can subsequently affect their attitudes and behaviours at work (Weiss & Cropanzano 1996). Also, research in the area of emotional intelligence has introduced a new appreciation of the role of emotional perception, understanding, and management in organisations, popularized in by Goleman (1995, 1998). More recently, there has been a proliferation of information on the topic of the role of affect in the work experience (Ashkanasy, 2003; Barsade Brief & Spataro, 2003; Brief & Weiss, 2003), indicating a renewed interest in the area of emotions and work.

70

Despite the renewed interest, little empirical research exists that specifically examines affective leader behaviours and their impacts on employee engagement. Some information can be derived from the engagement work of Kahn (1990), who found that supportive, resilient, and clarifying management, heightened psychological safety. Kahn found that much like supportive interpersonal relationships, supportive managerial environments allowed people to try and to fail without fear of the consequences. In addition, Louis (1986) found that leaders reinforce members' behaviours in ways that may create different degrees of supportiveness and openness (Louis, 1986). Studies by Laschinger et al. (2000) and Tan and Tan (2000) conclusively demonstrated that increased trust in management influenced employee’s levels of affective commitment. Furthermore, and again consistent with the role of affect on organisational outcomes, Connell et al. (2003) reported that trust in management, rather than directly influencing extra role performance (ERP), indirectly influenced ERP through its influence on affective commitment. Meta-analyses have also provided clear evidence in support of a relationship between organisational support and affective commitment. Rhoades and Eisenberger’s (2002) meta-analysis, based on 42 studies involving more than 11 000 employees, showed a strong relationship between employee perceptions of organisational support (POS) and affective commitment (AC). Similarly, Meyer et al.’s (2002) meta-analysis, combining the results of 18 studies involving 7128 participants, found a very strong predictive relationship between organisational support and affective commitment. Thus, leadership which is seen as supportive will influence engaged behaviour. Organisational support theory states that employees develop a perception over time, of how much their organisation values their input and cares for their

71

welfare (Stinglhamber & Vandenberghe, 2003). To the employee, a perception of being valued and cared for is a significant guarantee that help will be available, when needed, for the effective performance of one’s job and to cope with stressful situations (Stinglhamber & Vandenberghe, 2003). Also, support in the work environment is an important resource that enables employees to cope with or reduce the negative effects of various stressors within the work environment: The degree of support an individual receives in the workplace, from either colleagues or supervisors, may affect the process of coping with stress. For example, if an employee experiences conflict in a job but has a high level of support from peers or supervisors, the resulting anxiety or stress from the conflict may not occur (Carlson & Perrewe, 1999). Halbesleben (2006) similarly noted that support can widen an employee’s resources and restore or strengthen resources that may be lacking. Additionally, it has been suggested that employees who perceive a high level of support within the work environment are more likely to feel a responsibility to ‘pay back’ the organisation in terms of affective commitment (Wayne & Green, 1993). That is, the employee will have an emotional attachment to the organisation based on the perceived support. Various studies have shown that perceived organisational support can strengthen affective commitment to the organisation (O'Driscoll & Randall, 1999; Stinglhamber & Vandenberghe, 2003; Wayne & Green, 1993). Moreover, the literature suggests that social support is related to organisational commitment (Osca et al., 2005). For example Frone, Yardley, and Markel (1997) found that supervisor support was positively related to work commitment. Joiner and Bakalis (2006) found that employees who had positive

72

relationships with their co-workers and supervisors had a stronger commitment to their organisations. Eisenberger, Armeli, and Rexwinkel (2001) concluded that organisational support had a direct and positive influence on affective commitment. It was also found that affective commitment was associated with a supportive and caring supervisor (Rhoades, Eisenberger, & Armeli, 2001). In addition, Iverson and Buttigieg (1999) found that colleague support had a positive effect on affective commitment. Thomas, Bliese, and Jex (2005) found that supervisory support had a positive relationship with both affective and continuance commitment. Finally, Joiner and Bakalis (2006) found that both colleague and supervisor support were associated with higher levels of affective commitment. In summary, the role of affect on organisational outcomes has certainly been articulated in various forms of research. What is lacking currently is a specific focus on leader behaviours that act to engender engaged responses in employees. Transformational leadership as a construct has in part addressed some of these responses, as will be discussed in further detail in the next section. Transformational Leadership In their model of employee engagement, Macey and Schneider (2008) show where transformational leadership literature provides examples of outcomes of transformational leadership that closely align with both affective and behavioural engagement. Some of the outcomes of transformational leadership include an investment of identity in the organisation which results in a passion for work and the capacity to think independently (Bass & Avolio, 1990); the capacity to think autonomously and challenge convention as necessary for the efficacy of the organisation (Divir, Eden, Avolio & Shamir, 2002). The overriding message here is

73

that cognitive, affective and behavioural engagement are more achievable under certain conditions, and that the leadership under which they work is critical to their levels of engagement (Macey & Schneider, 2008). Another important (and closely aligned) antecedent highlighted by Macey and Schneider, includes the notion of trust. In their model, they specifically propose that feelings of trust mediate the relationship between leadership behaviour and behavioural engagement (see figure 1). The term ‘Transformational Leadership’ was first articulated by James V. Downton in 1973 in Rebel Leadership: Commitment and Charisma in a Revolutionary Process. James Burns (1978) later articulated the concepts of transformational and transactional leadership in his treatment of political leadership to describe a situation in which the leader empowers workers to achieve an articulated vision of the organisation and its mission, leading to increases in productivity, employee morale, and job satisfaction as well as heightened personal and professional growth. In order to measure these types of leadership styles, Bass and Avolio (1990) developed the Multifactor Leadership Questionnaire (MLQ). This instrument was designed to measure a full range of leadership styles and behaviours through four transformational components (idealised influence; Inspirational motivation; Intellectual stimulation; Individualised consideration) and two transactional components: contingent reward and management by exception (Bass and Avolio 1990). Transformational leadership is a well regarded and utilized leadership construct. In their review of the leadership literature over a 10 year period, Judge and Bono (2000) concluded that “it is safe to say that transformational leadership theory

74

has garnered most of the attention in recent leadership research” (p.751), and has provided significant research linking transformational leader behaviours with organisational outcomes (Chen 2005; Lowe; Schaubroeck, Lam & Cha, 2007). A meta-analysis conducted by Lowe, Kroeck, and Sivasubramaniam (1996) confirmed significant correlations between the MLQ transformational components, with: effectiveness; satisfaction; and extra effort (Bass, 1997). Overall, the MLQ can be regarded as providing a useful model for assessing leadership comprised of both affective and cognitive components. However, a criticism of the MLQ was the fact that much of the formative construct research was based on 'distant' leaders as opposed to 'immediate/nearby' leaders, with the former being described as having a more ideological orientation and a strong sense of mission, which they expressed with little concern for personal criticism or sanction (Shamir, 1995). In contrast, 'nearby' or immediate leaders were seen as, sociable, open and considerate of others, having a sense of humour, having a high level of expertise in their field, and being dynamic and active. Shamir (1995) also found that in close (immediate) leadership situations, the charismatic effects of the leader depend more strongly on the leader’s observable qualities and behaviours that produce identification with the leader and empowerment by the leader. Transactional and supportive behaviours that generated trust in the leader were also more relevant in close leadership situations (Shamir, 1995). The distinction between distant and nearby approaches is also consistent with Yukl’s (1999) distinction between the “heroic” model of the charismatic leader, with whom followers identify strongly, and may become passionately devoted to, versus the charismatic leader who effects a process of influence encouraging followers to internalise what is being advocated, by linking the task objectives, and the mission, to the followers’ core

75

values and self-identity. Many of these characteristics have much in common with LOS, as discussed earlier. An investigation of transformational leadership in the UK public sector, based on eliciting constructs of leadership from staff in relation to behaviour of their immediate line manager, resulted in the emergence of a different understanding of the nature of leadership from the “heroic” models of leadership (Alimo-Metcalfe & Alban-Metcalfe, 2001, 2002, 2005), and one which is more in line with Greenleaf’s (1970) notion of “servant” leadership which emphasizes collaboration, trust, empathy, and the ethical use of power. The premise of this latter theory is basically in its supportive thrust - that the leader is a servant first, making the conscious decision to lead in order to better serve others, not to increase their own power. The objective of this style of leadership is to enhance the growth of individuals in the organisation and increase teamwork and personal involvement (Shamir, 1995). The result of Alimo-Metcalfe and Alban-Metcalfe’s (2000, 2005) research into the UK public sector, was the development of a new Transformational leadership questionnaire (TLQ), which took into account the notion of servant leadership as discussed. The instruments that emerged from the research, the TLQ (Public) and the TLQ (Private sector version), have both been shown to achieve the required psychometric properties of reliability and content, construct, and convergent validity (Alban-Metcalfe, & Alimo-Metcalfe, 2001, 2005). It is interesting to note that the authors found very little difference between the findings from public sector organisations and from individuals in various industries in the private sector. The most conspicuous difference with the North American research (e.g. Bass, 1990) was

76

in relation to the concept of ‘close’/ ‘nearby’ leaders, which revealed more of a “leader as servant” notion of leadership (Greenleaf, 1970) as discussed earlier. The distinction between ‘distant’ and ‘close/nearby’ leadership is important. While the perception of leaders from different organisational levels has been established as significantly different, the affective influence of the immediate supervisor in shaping engaging behaviours of subordinates is also likely to be significantly different to that of senior organisational leaders. Thus it is another agenda of this research to elicit perceptions of leaders from their immediate subordinates, the level of relationship considered to have the most affective potency (Taylor, 2004). In summary, in accordance with Macey and Schneider’s (2008) arguments for consideration of transformational leadership as an important antecedent to engagement, it is of interest to asses this dynamic for the purposes of this research. Specifically, this study is interested in examining affective leader behaviours, and their impact on subordinates’ levels of cognitive, affective and behavioural engagement. The affective elements of the transformational leadership construct in particular, will be used to examine affective leader behaviours. Finally, this research is restricted to the study of immediate or nearby leaders as this level of leadership is considered of particular importance in terms of its impact on employee engagement. Terms used interchangeably in this literature review, include: nearby, close leaders and immediate leaders/supervisors.

77

Scope, summary and hypotheses The overall aim of this study is to test a model defining the various interrelationships between the constructs of leadership behaviours, and cognitive, affective and behavioural engagement. This final section of this chapter will provide a summary of literature reviewed, with the aim of articulating and justifying the proposed model. In addition, it will discuss briefly, some of the measures to be used, in order to test the proposed model. Elements of pre-existing constructs that have established psychometric validity and reliability, have been used to test the proposed model. Accordingly, the justifications for their use are discussed so that the reader is able to make sense of the proposed model (please refer to Chapter III for a full discussion of the various measures used in this research). As discussed, limited empirical research exists that specifically examines leader behaviours and their impacts on employee engagement (i.e., cognitive, affective and behavioural). The aim of this study is to begin to address this gap. For the purposes of this research, employee engagement is defined as “an employee’s attitudinal attachment to his or her job or company, intention to act in the company’s best interest and willingness to invest discretionary effort in achieving business goals” (Jesuthalen, 2003, p. 57). As discussed, engagement for this thesis is comprised of cognitive engagement, affective engagement and behavioural engagement (Macey & Schneider, 2008), with the process being hypothesized as directional. Each of the components of the model shown below will be briefly summarized in the following section.

78

F Figure 4: Engagement En sequence E Engaging L Leader Behaaviours O particulaar interest too this researrch is an examination of Of o the behavviours and characteeristics of managers m whho achieve high h levels of engagem ment in theirr staff. While thhe notion off engaging leadership l i a new areea of researcch, many arreas of is empirical leadershiip research overlap, o moost notably within w the transformati t ional o affect on organisatioonal behavioour. This leadershhip literaturre, and withiin the role of researchh has arguedd that affecttive leader behaviours b are the mosst significannt drivers of the coognitive, afffective and behaviouraal engagemeent of emplooyees. F Focusing sppecifically on o the leadeership literatture, George (2000) arggued that the affective or emotional dim mensions asssociated witth leadershipp have not bbeen sufficienntly researcched or com mmented upoon. Althoug gh it is wideely recognizzed that transforrmational leeaders engennder an affeective response in emplloyees, veryy little researchh has been conducted c o the conseequences off how emplooyees feel abbout their on designaated leaders.. Alimo-Meetcalfe’s (20001) Transfo ormational Leadership L Questioonaire (TLQ Q) was desiggned to empphasize the importance i of followerrs’ feelings toward their t leaderrs and will be b used as thhe basis for assessing leader behavviours in this reseearch.

79

Cognitive Engagement Prior research on meaningful work (Hackman & Oldham, 1975), role clarity (Jackson & Schuler, 1985), and perceived fit (Cable & Judge, 1996) suggests that employees seek meaning in their jobs by wanting to see how they fit with or contribute to their organisation. Cognitive engagement for the purposes of this research is essentially a dimension involving processes which lead to an employee’s understanding of their role, perceived fit and the overall strategic imperatives of their job, with the result being cognitive engagement. As mentioned previously, the “activation” of cognitive engagement is highly dependent on the communication strategies/behaviours of the sender. Thus, again, this highlights the critical role of immediate supervisors/leaders as the conduits or enablers of cognitive engagement, as they are effectively in control of the many aspects of communication that provides the transparency and two way dialogs between individual capability and business growth strategies. Put another way, the conduit of cognitive engagement exists predominantly in the capabilities and behaviours of direct supervisors. Thus, for the purposes of this study, cognitive engagement will be measured by Sims et al.’s (1976) job characteristics inventory (feedback subscale), Sawyers (1992) goal clarity measure, and Boswell’s (2003) Line of Sight measure. Affective Engagement Affective engagement is a broad multi-dimensional construct which contains the affective and energic/arousal components of a group of related constructs focused on individuals’ relationships with their work roles (Maslach & Leiter, 2008; Salanova, Agut, & Peira, 2005). It overlaps conceptually with organisational commitment (Macey & Schneider, 2008; Meyer & Allen, 1997; Porter, Steers,

80

Mowday, & Boulian, 1974; Wellins & Concelman, 2005); job involvement (Lawler & Hall, 1970; Lodahl & Kejner, 1965); and job satisfaction (Burke, 2008; Harter et al., 2002; Kanungo, 1982). Consequently, affective engagement will be measured by the adaptive, energic, affective components of the contributing constructs. It is of particular interest in this research to establish the nature of affective engagement, to determine its foci (direct leader or organisational), and to determine its antecedents (cognitive or leader). As discussed previously in the section on leadership, research in both the academic and applied fields has indicated the significance of the relationships between direct leaders and work outcomes (O'Driscoll & Randall, 1999; Stinglhamber & Vandenberghe, 2003; Wayne & Green, 1993); leaders and trust (Laschinger et al., 2000; Tan & Tan, 2000); and leadership with follower thinking, feeling and behaviour (Albrecht, 2005). It is therefore proposed that affective engagement for this research has a two part foci. Initially, affective engagement is hypothesized to be manifested as feelings of commitment towards a supervisor / direct leader, and subsequently, affective commitment to the organisation as a whole. Additionally, it is of interest in this research to establish the nature of affective engagement, to determine whether it is a consequence of cognitive engagement (as hypothesised), or a direct consequence of feeling towards ones direct leader or supervisor. The constructs identified that currently best serve this purpose are Allen and Meyers (1997) organisational commitment scale (measuring affective commitment); and Becker et al.’s (1996) Job satisfaction (supervisory commitment subscale) which measures satisfaction and identification with ones supervisor.

81

As an aside, because the continuance commitment subscale of Allen and Meyers (1997) organisational commitment scale is frequently used in conjunction with other measures to assess engagement (Corporate Leadership Council, 2004), justification for its absence should be noted. Continuance organisational commitment refers to the cost an employee will pay for leaving the organisation compared with the cost of staying within the organisation (Allen & Meyer, 1990). Thus, when an employee has continuance commitment the cost of staying with the organisation outweighs the cost of leaving e.g., “My salary is too good to leave the organisation”. Using the argument provided previously, namely that the measurement of engagement should be restricted to the affective and energic/arousal components of pre-existing constructs (Macey & Schneider, 2008), then the continuance commitment subscale is unlikely to measure engagement. Instead, it is far more likely to tap a construct outside that of engagement, notably, a rational consideration based on the degree and/or number of investments employees make, and on the perceived shortage of alternative choices (Allen & Meyer 1990). It will consequently not be used in this research. Finally, as indicated previously, multi focus commitment research literature has discussed the need to distinguish between commitments to different workplace targets (Ellemers, De Gilder, & Van den Heuvel, 1998). For instance, Riketta and Van Dick (2005) found that closer (lower order) levels of attachment (i.e. direct supervisory) were generally stronger than attachment to more distant, higher order targets of attachment (i.e. organisation). Similarly, in research by Billings, Eveleth, and Gilbert, (1996), the supervisor was shown to be a closer and more potent focus of commitment than the organisation. Consequently, the Becker et al. (1996) supervisor- related commitment measure was used to assess the role that

82

commitment to a supervisor might play in the engagement process. This argument is based on the premise that when employees care about and feel involvement and identification with a supervisor, they will be more likely to embrace the organisation’s goals and invest effort to attain them. As will be shown in the proposed model (Figure 5), supervisory commitment will be hypothesised to predict affective commitment, thus reflecting the differential impacts of commitment foci as aspects of affective engagement. Behavioural Engagement. Macey and Schneider (2008) defined the construct of behavioural engagement, as ‘adaptive behaviour intended to serve an organisational purpose” (p.18). This construct of engagement is seen as directly observable behaviour in the work context and which includes innovative behaviours, initiative, proactive behaviours and going above and beyond what might otherwise be expected. Potentially contributing constructs include; that of Organisational Citizenship Behaviour (OCB) and its variants; role expansion (Crant, 2000) and personal initiative (Frese & Fay, 2001) and, turnover intentions (Hom, Griffeth, & Sellaro, 1984). For the purposes of this research, discretionary effort and intent to turnover will be used as the basis for measurement. Intent to Stay. As discussed, research findings attest to significant inefficiencies and costs associated with employee turnover. These costs are usually measured in terms of loss of experience and expertise, and the human resourcing costs associated with time and money required to select and train replacement staff (Tziner & Birati, 1996). Although empirically determined associations between turnover intention and actual turnover have typically been found to be modest

83

(Kirschenbaum & Weisberg, 1990), it is arguably more practical for organisations to manage turnover intention rather than actual turnover, since intentions can be modified through organisational intervention. Once employees have left the organisation, the loss has already been incurred. Discretionary effort. Extra Role Performance (ERP) or discretionary effort is being increasingly recognised as an important dimension of organisational performance (Borman & Motowidlo, 1993; LePine et al., 2002) and as an indicator of behavioural engagement (Kahn,1990; Towers Perrin, 2003). Borman and Motowidlo (1997, p.100) argued that while task performance is best understood in terms of activities that either support or directly contribute to the transformation of the organisation’s inputs to output, discretionary effort (ERP) is best understood in terms of “activities that support the social and psychological context in which the organisation’s technical core is embedded”. In support of the importance of discretionary effort, Borman and Brush (1993) found that 55% of overall job performance could reliably be identified as task performance and 30% as discretionary effort.

Research Model The specific aim of this study is to test a model defining the various interrelationships between the constructs of affective leader behaviours, and the engagement dimensions of cognitive, affective and behavioural engagement. The proposed model is shown in Figure 5.

84

F Figure 5: Prroposed moodel of emplloyee engag gement (Note: foor the simpliciity of presentaation, only thee proposed fullly mediated reelationships are shown).

Hypotheeses T (Engagging) Transfformational Leadership The p Questionnnaire (Publicc Sector Versionn) was basedd on researcch conducteed with man nagers and professional p ls, at all levels (B Board to miiddle level),, in the NHS S and local governmennt in the UK K. The scale itsself consistss of 6 factorrs, notably: Showing geenuine conccern; Enabliing; Being Honest and Consisttent; Being Accessible; Networkin ng and Achiieving; and Being r is specifically s interested in i the role of o affect drivven Decisivve. As this research leadershhip behaviours, only thhe first four of the abov ve –mentionned factors ((those reflectinng affect) will w be used. F this research, it is proposed For p is that social and emotioonal supportt provided by imm mediate supeervisors, is liikely to preedict the threee elementss of engagem ment, specificcally: cognittive; affectivve; and behhavioural en ngagement. H1

A Affective L Leader Behaaviours (meaasured as: Showing S Geenuine Conccern; E Enabling annd Empoweering; Beingg Honest and d Consistennt; Being Oppen and A Accessible) ), will have a positive relationship r with:

85

(a) Cognitive Engagement (b) Supervisory Commitment (c) Affective Commitment (d) Discretionary effort (e) Intent to Stay

H2:

Cognitive engagement, will have a positive relationship with: (a) Supervisory Commitment (b) Affective Commitment (c) Discretionary Effort (d) Intent to Stay

H3:

Supervisory commitment will have a positive relationship with (a) Affective Commitment (b) Discretionary Effort (c) Intent to stay

H4:

Affective commitment will have a positive relationship with (a) Discretionary Effort (b) Intent to Stay

86

Mediateed Relationsships F Figure 6 preesents a meediation moddel as devissed by Baronn and Kennny (1986). This moodel indicates that a varriable can inntervene beetween a stim mulus (behaaviour) and a reesponse (inttending to leeave the orgganisation), which is “tthe most genneric formulaation of a mediation hyppothesis” (B Baron & Keenny, 1986, p. 1176). O Overall, a variablee is a mediator to the exxtent that it accounts fo or the relatioonship betw ween a predictoor and a critterion variabble (Baron & Kenny, 1986). This study exploored to what exxtent cognitiive engagem ment, and thhe affective engagemennt measures of affectivve and superrvisory com mmitment acccount for th he relationshhip betweenn affective leader behaviours b a the behaavioural enggagement in and ndicators off discretionaary effort and inteent to stay. F Figure 6: Baron B and Kenny's K meddiation mod del

M Mediating e effects of Coognitive Enggagement B Because as argued in thhe literaturee, cognitive engagemennt can be reggarded as a pre - affective a staate, this reseearch is inteerested in deetermining the t role of ccognitive engagem ment in the context of multiple m relationships. Therefore the followinng hypotheeses are propposed.

87

H5:

Cognitive Engagement will mediate the relationships between affective leader behaviours, and (a) Affective Commitment (b) Supervisory Commitment (c) Intent to Stay (d) Discretionary Effort

Mediating effects of Affective Commitment The proposed associations between affective commitment and discretionary effort and turnover intentions, are based on the premise that employees who feel an emotional attachment to their organisation will likely demonstrate a willingness to invest discretionary effort (Bishop, Scott & Burroughs, 2000), and be less likely than others to exit their organisation (Meyer & Allen, 1997; Price & Mueller, 1986). Hypothesis 6 reflects cognitive engagement influencing supervisory commitment, which in turn influences affective commitment, which in turn influences intent to stay and discretionary effort. The following hypotheses are proposed: H6:

Supervisory commitment will mediate the relationships between: (a) cognitive engagement and affective commitment (b) cognitive engagement and intent to stay (c) cognitive engagement and discretionary effort

H7:

Affective commitment will mediate the relationships between:

88

(a) supervisory commitment and intent to stay; and (a) supervisory commitment and discretionary effort.

Summary This study is aimed at examining engaging leader behaviours, their impact on cognitive, affective and behavioural engagement variables. Researchers (e.g., Cropanzano & Mitchell, 2005; Moorman et al., 1998) have called for more research to be directed toward modelling and testing set of relationships which better approximate the complexity associated with contemporary organisational experience. This research attempts to extend the existing literature by testing, using structural equation modelling, the set of relationships as hypothesised above. As demonstrated in this chapter, employee engagement has a strong normative base, particularly in the applied literature, as used and leveraged by consulting firms. Less has been achieved in the academic literature and as a result, engagement has a shallow research base from which to work. The aim of this study is to address this gap. The model of engagement as outlined in this chapter places employee engagement as a connecting process comprised of cognitive, affective and behavioural engagement. It specifically provides an individual level of analysis. In particular, it is interested in addressing the effects of affect driven leadership behaviours on individuals’ levels of cognitive, affective and behavioural engagement. In addition, it reflects the notion that affective engagement is expressed at different levels of foci. Defined as such, there are several elements which can be

89

researched through established measures. The next chapter deals with how these measures will be applied, in the context of the overall research methodology and design.

90

Chapter III: Method and Results In the previous chapter, the literature on engagement was reviewed. This involved an assessment of both the academic as well as the practitioner oriented literature. Macey and Schneider’s (2008) model of engagement was used as a basis for the examination of engagement, as was Kahn’s (2001) model of cognitive, affective and behavioural engagement. The latter models were used as the basis of a model for this current research. In addition, the previous chapter discussed the related engagement constructs (i.e., organisational commitment, job satisfaction), and discussed their contribution, degree of overlap and where they separate as individual constructs. Finally, a model for the research was presented and hypotheses were proposed (see Chapter II, p. 83 – 87). This chapter provides the results of a test of the hypotheses. First, the research design utilized for the research is overviewed and the participants, procedures and statistical applications used in the research will be discussed. Second, the measurement model will be established, that is an examination of the validity and reliability of the dimensions of engagement (cognitive, affective and behavioural), as well as affective leader behaviours via confirmatory factor analysis will be completed. Third, an analysis of the structural model will be conducted. Specifically, the proposed mediation model using structural equation modeling will be examined. Finally, the findings in context of previous research and the current research hypotheses will be discussed.

91

Method Sample Letters were sent to Director Generals of seven Queensland State Government Departments in Brisbane asking them to participate in organisation wide engaging leadership research. The departments were selected on the basis of existing relationships with organisational leaders, who had expressed their willingness to potentially participate in the research. The researcher offered to deliver a report at the conclusion of the research, with key overall findings. Five organisations expressed an interest in participating, with all organisations proceeding to participate in the research as follows: Department of Local Government, Sport and Recreation (DLGSR); Department of Natural Resources and Water (NRW); Environmental Protection Agency (EPA); Forestry Plantations Queensland Office (FPQ); and, Queensland Audit Office (QAO). The organisations that participated in the study differed significantly in terms of their core business activities, with a diverse range of professions such as sport and recreation activities, environmental protection, financial and audit responsibilities, local government regulation and so on. In addition, the questionnaires were released in each of the departments, on an organisation wide basis, though 70 operational forestry workers in Forestry Plantations Queensland were excluded on the basis of extremely limited access to computers. In summary, overall there was considerable heterogeneity within the sample, both in terms of departmental core business, the organisational level of respondents and jobs.

92

Participants The survey instrument (see below) was accessed via an email forwarded to all organisation staff by the department’s director general, or representative. Thus all organisational staff had the opportunity to participate. Figure 7 provides a summary of the number of respondents from each participating organisation. Figure 7: Participating Organisations

NRW

DLGSR

EPA

FPQ

QAO

TOTAL

Survey responses

333

78

177

27

14

635

Percentage of used responses

(52.4%)

(12.3%)

(27.9%)

(4.3%)

(2.2%)

(100%)

Note abbreviations as follows: Department of Local Government, Sport and Recreation (DLGSR); Department of Natural Resources and Water (NRW); Environmental Protection Agency (EPA); Forestry Plantations Queensland Office (FPQ); and, Queensland Audit Office (QAO)

Nine hundred and sixty-nine people out of a workforce of approximately 2430 people (40%) participated in the survey and completed the demographic information. Of those, 635 completed the survey in a useable form. The remaining 334 cases did not complete more than two thirds of the survey and were thus excluded from the analysis. An examination of those deleted did not yield any particular pattern, in terms of demographics, departments or position levels or gender.

93

The average age range of participants in the sample was in the 40-49 yrs age range which represented 25.9% of the sample. The average tenure in the organisation was 11.54 years. The average tenure in a current position was 3.18 years. Females comprised 48.5% (308) of the sample and 51.5% (327) were male. Procedure The data were collected by way of a questionnaire accessed through survey monkey. Survey monkey is an online questionnaire tool where the researcher has the capacity to create a survey and have it administered through survey monkey’s online data collection tools. The advantage of collecting surveys in this manner are that the surveys can be immediately forwarded to staff without the cost or time impost of regular postage, and the results obtained almost instantaneously (Cavana, Delahaye, & Sekaran, 2001). On line surveys are also interactive, with check boxes and pull down menus, therefore making the survey quicker and more expedient to complete (Sharp et al., 2002). A negative of this method includes the restriction of staff that do not have access to a computer. However, as all of the staff targeted for this survey had access to computer facilities, this was not considered to be a problem. Other potential negatives of this method include accessing staff who are not comfortable using computers (Newman, 2007). However, as the public sector employees are all independently allocated computers for their work, and use such as the primary tool for completing their work tasks, this was not considered to be an issue for the integrity of the sample (Cavana et al., 2001). Once research approval from the CEO’s was gained, an email copy of the questionnaire and covering letter was forwarded to the CEOs of the participating organisations for consideration and approval prior to release. Only very minor

94

alterations in the covering letter pertaining to different terminology used by various departments were made. Once the research had clearance to proceed, the CEO was once again emailed a link to the survey via a covering email, which was then on forwarded to all staff. The covering page detailed what the study was about, who was doing the study, the rationale for the study, confidentiality and anonymity, what was required of them and an offer to supply a summary of results from the research. Participants were also informed on the cover sheet that their participation was completely voluntary, and that if they agreed to participate, that they could withdraw from participation at any time during the project without comment or penalty. Participants were informed that their participation would involve completing a questionnaire which asked questions regarding their perceptions of their supervisor’s leadership behaviours, as well as some questions concerning their attitudes regarding work. Employees were then asked to access a link provided, which took them directly into the survey. The questionnaire (in prior pilot trials) took approximately 20 minutes to complete. After completion of the survey, survey monkey automatically electronically “collected” the survey, for immediate access by the researcher. Within the survey itself, the questionnaire asked for a number of demographic details. In addition, it asked for the voluntary disclosure of participants and their supervisor’s contact details. This latter information - with its obvious potential ethical issues (Neuman, 2007) was requested for future research purposes, in which follow up information in the form of interviews will be requested where groups of staff have identified an exceptional leader. This further research is intended to elicit more detailed qualitative information (with semi-structured

95

interviews) regarding the behavioural characteristics of ‘engaging’ leaders as rated quantitatively by staff. In disclosing the information, participants were further assured that no persons outside the research team would have access to the data obtained during the project. The final information from the research would be made available, but as indicated earlier, would contain no identifiable information, either about individual respondents, supervisors, or the individual departments themselves. In short, other than the fact that the data pertained to the Queensland Public sector, no other identifiable characteristics would be provided in any published works. The Research and Ethics Committee of the Queensland University of Technology provided ethical approval for this study. An executive summary of the research findings was given to participating organisations on completion of the study with a letter attached for distribution to all staff. Measures The engaging leadership questionnaire was a composite survey comprised of demographic questions as well as aspects of seven pre-existing questionnaires, selected to measure the components of cognitive, affective and behavioural engagement. A copy of the questionnaire is provided as Appendix 2. The research version of the MLQ (public) (Alban-Metcalf & Alimo –Metcalf, 2002) was used to assess leader behaviours. Table 1 shows the list of the instruments used, each of which will be discussed in greater detail thereafter. The internal reliability (Cronbach alphas) for each scale is shown in Table 6.

96

Table 1: Summary of Measures used

Construct to be measured

Questionnaire

Source

Engaging Leader Behaviours

Transformational Engaging leadership Questionnaire (TLQ) public

Alimo-Metcalf & Alban-Metcalfe (2002)

Cognitive Engagement

1. Goal Clarity 2. Communication

1. Goal and Process Clarity. Sawyer (1992) 2. Job Satisfaction Survey. (communication

satisfaction Affective Engagement

Behavioural engagement

sub scale) Spector (1985)

1. Affective Commitment

1. Meyer & Allen (1990). Affective, Normative

scale 2. Supervisor Commitment

2. Supervisor-Related Commitment. Becker,

1. Discretionary effort

1. Organisational Citizenship Behaviour.

(altruism) 2. Discretionary effort (conscientiousness) 3. Intent to Turnover

and Continuance Commitment. Eveleth & Gilbert (1996)

Smith, Organ & Near (1983)

2. Chinese Organisational Citizenship 3.

Behaviour. Farh, Earley & Lin (1997) Michigan Organisational Assessment Questionnaire. Cammann, Fishman, jenkins & Klesh, 1979; (intent to turnover subscale)

Leader Behaviours (Transformational Leadership Questionnaire) The Transformational Leadership Questionnaire (TLQ - Public) identifies and assesses individual leadership behaviours. This tool was designed for professional staff and middle to senior managers in the public sector. Table 2 shows the internal consistency (Cronbach’s α coefficient) and inter-item correlation coefficients for TLQ public. The results were based on 2013 male and female National Health

97

Service managers and professionals (Alban-Metcalfe & Alimo-Metcalfe, 2000). The sample was subordinates rating their current or a previous Line Manager.

Table 2: Internal consistency and correlation coefficients for TLQ (public)

Scale

Number of items

Cronbach’s α coefficient

Inter-item correlations

Mean

Standard deviation

Coefficient of variation

1 – Showing Genuine Concern = Genuine concern for other’s well-being & development

6

.91

.56 - .73

4.18

1.10

26.3

2 – Networking & Achieving = Inspirational communicator, networker & achiever

6

.85

.42 - .55

4.49

0.94

20.9

3 – Enabling = Empowers, delegates, develops potential

6

.86

.39 - .62

4.64

1.03

22.2

4 – Being Honest & Transparent = Transparency: Honesty & consistency

4

.81

.47 - .58

4.26

1.16

27.2

5 – Acting with Integrity = Integrity & openness to ideas and advice

6

.85

.43 - .59

4.37

0.99

22.7

6 – Being Accessible = Accessible, approachable, in-touch

5

.82

.42 - .53

4.58

1.01

22.1

The initial research on the TLQ by Alimo-Metcalfe and Alban-Metcalfe (2001) found two higher order transformational factors with eigenvalues greater than one. The two 'higher-order' factors were interpreted as: Internal-orientation which consisted of social – emotional support items (Factor I) and External-orientation (Factor II). The two factors were negatively correlated, r = -.50. Factor I (accounting for 42.9% of the variance) comprised: Genuine concern for others; Empowers, develops potential; Integrity, trustworthy, honest and open; Accessibility,

98

approachability; Encourages critical and strategic thinking (negative). Factor 2 (accounting for 12.3% of the variance) comprised: Inspirational networker and promoter, Decisiveness, determination, self-confidence (negative); Political sensitivity and skills. As the nature of affective leader behaviours was of specific interest in this research, only the affective or social – emotional support factors were used in this research (scales 1, 4, 5 and 6 respectively – see Table 4). Cognitive Engagement Cognitive engagement was measured for this research by three measures measuring respectively, goal and process clarity, communication satisfaction and connection. Goal and process clarity was measured using Sawyer’s (1992) goal clarity subscale, which is designed to measure clarity about the jobs outcome goals and objectives, for example, respondents are asked to indicate the degree of certainty they feel about “ how my work relates to the overall objectives of my work unit”. Respondents were asked to respond using a Likert type scale where 1= very uncertain, and 6 = very certain. The alpha coefficient for this scale was .92 (Sawyer, 1992). The alpha coefficient for the current research was also .92. Communication satisfaction was measured using Spector’s (1985) job satisfaction measure which measures satisfaction of nine job facets. The communication satisfaction subscale was used for this research. Examples of items include: “The goals of this organisation are not clear to me”, and “I often do not know what is going on with the organisation”. Both of these questions were reverse

99

scored. The alpha coefficient for this scale was .89 (Blau, 1999). The alpha coefficient for the current research was .93. Connection (Line of Sight) was measured using an item from Rucci’s (2007) human Capital Analytics survey (Rucci, 2007). The survey was designed to assess line of sight with the organisation’s objectives. An example of the item was “I understand exactly how my job contributes to the goals and objectives of my organisation”. Respondents were asked to respond using a Likert type scale where 1= very uncertain, and 6 = very certain. The alpha coefficient for this scale was .84 (Rucci, 2007) Affective engagement. Affective Engagement for this research was measured by both affective commitment and supervisory commitment. These measures were selected to capture both the immediate foci and organisational foci of affective engagement. Affective commitment. Organisational commitment is commonly defined as employees’ interest in, and connection to, an organisation (Hunt et al., 1989; Meyer & Allen, 1997). Meyer and Allen (1984) initially proposed that a distinction be made between Affective Commitment (AC) and Continuance Commitment (CC), with AC denoting an emotional attachment to, and involvement in, the organisation and CC denoting the perceived costs associated with leaving the organisation. Allen and Meyer (1990) later suggested the third discrete component, termed Normative Commitment (NC), which reflects a perceived obligation to remain in the organisation.

100

As neither the continuance commitment scale nor the Normative commitment scale were considered to reflect the affective, engergic, adaptive components of engagement, the latter two scales were not used. Only the Affective commitment scale was used in this research. Affective organisational commitment was assessed using eight items, for example, “I would be happy to spend the rest of my career with this organisation”. Four of the eight affective commitment items were reverse scored as the original four items reflected low levels of affective commitment. The affective commitment items were measured on a six-point scale ranging from 1 = strongly disagree to 6 = strongly agree. The Cronbach’s alpha for affective commitment, ranged from .77 to .88 (Hunt et al., 1989; Meyer & Allen, 1997). The alpha coefficient for the current research was .85. Supervisor Related Commitment. As this research was interested in leader behaviour and various domains of engagement, there was interest in determining the foci of the affective commitment i.e., whether behaviours creating engagement, whether that engagement in turn was directed back at the leader, or whether it was further extendable to the organisation. Of particular interest to the research was to differentiate between the foci of affective engagement was to the supervisor or to the organisation. To that end, this research used the ‘Supervisor – Related commitment scale’ developed by Becker, Billings, Eveleth and Gilbert (1996) which measures identification with a supervisor. A confirmatory factor analysis suggested that commitment to the organisation and the supervisor were empirically distinct. Responses were obtained using a 7–point Likert type scale where 1= strongly disagree and 7= strongly agree. A sample question from the measure was “When someone criticises my supervisor, it feels like a personal insult”. Coefficient alpha

101

for supervisor-related commitment based on identification was α= 0.85 (Becker et al., 1996). The alpha coefficient for the current research was .91.

Behavioural Engagement Behavioural Engagement for this research was measured by measures of Turnover intention, and discretionary effort. Intention to stay. Turnover intention was assessed using four items from the Michigan Organisational Assessment Questionnaire (Intention to Turnover) subscale, by Cammann, Fichman, Jenkins and Klesh (1979). Participants used a six-point scale ranging from 1 = strongly disagree to 6 = strongly agree. The participants were asked to indicate the likelihood that they would leave the organisation in the near future, for example, “It is very possible that I will look for a new job soon”. Reported Cronbach’s alphas for turnover intentions range between .67 to .95 (Fields, 2002). The alpha coefficient for the current research was .86. Discretionary effort. Discretionary effort was assessed using the subscales of altruism from the Organizational Citizenship Behaviour Measure (OCB) developed by Smith, Organ and Near (1983); and the conscientiousness subscale of the Organizational Citizenship Scale by Fahr, Earley and Lin (1997). The alpha coefficient for the current research was .65. The OCB measure (Smith, Organ & Near, 1983) uses 16 items to describe two dimensions of organisational behaviour. The two dimensions are altruism and generalised compliance. Altruism is defined as helping co-workers personally, such as making innovative suggestions to improve the department. Generalised compliance is impersonal helpful behaviour such as punctuality and not wasting time

102

on the job. In line with the tenant of choosing measures that reflect adaptivity, the altruism subscale only was selected for inclusion. Coefficient alphas for altruism ranged from .86 to .91 (Cropanzano et al., 1997). Responses were obtained using a 7 point Likert-type scale where 1 = never and 7 = always. The Chinese Organizational Citizenship Scale was developed by Fahr, Earley and Lin (1997). This measure uses 20 items to describe five dimensions of organizational citizenship behaviour, namely: identification with the company; altruism; conscientiousness; interpersonal harmony; and protecting company resources. The conscientiousness items were considered of interest for this research. This particular subscale measured items of discretionary effort reflecting behavioural engagement. Examples of items were; “Does not mind taking on new or challenging assignments” or “complies with company rules and procedures even when nobody watches and no evidence can be traced”. Coefficient alpha for the contentiousness subscale was .82. Responses were obtained using a 7-point Likert scale where 1 = strongly disagree and 7 = strongly agree.

Results Data Screening The assumptions of multivariate normality and linearity were evaluated through SPSS. The data were screened for the presence of outliers. A score was considered to be an outlier if it was three or more standard deviations above or below the mean (Tabachnick & Fidell, 1996). Results were run with and without outliers removed and no difference in findings were observed. Two univariate

103

outliers were removed as they were scores that exceeded item response scales. Other univariate outliers were retained as they were deemed to be part of the population. Data screening for respondent errors and omissions was conducted prior to analysis. There were up to 86 cases with missing data on some items. Dialla (2000) recommended that listwise deletion is the safest method of dealing with missing data when the data are missing randomly, are no more than 10% of the sample, and the sample remains large after listwise deletion. Although the data was randomly missing it did exceed more than 10% of the sample. If listwise deletion was applied then the sample would have been reduced to 293 cases (down from 635 cases). Since the sample was relatively large and no pattern was observed in respect of the missing data, EM (Expectation Maximisation) algorithm was chosen as the most reasonable approach to dealing with the missing data. The EM algorithm is a two-step iterative procedure whereby missing values are imputed, and a covariance matrix and mean vector are subsequently estimated. In almost all cases, the use of an EM-generated covariance matrix results in more accurate estimates of coefficient alpha (Enders, 2003). This is on the basis that it provides the expectation of the missing data from both the observed values and the current estimate of the parameters – which are then substituted for the missing data, enabling the estimation of the underlying population parameters with minimal bias and error (McKnight, 2007).

104

Measurement Model Higher order cognitive engagement scale It was of interest to test for a higher order factor (cognitive engagement) to determine if this provided a better fit for the data. Consequently, AMOS 16 (Byrne, 2001) was used to compare the fit of three nested models ranging from a singlefactor model (a) incorporating goal clarity (GC), job feedback (JF) and connection (C); a two factor model incorporating GC and JF (factor 1) and C (factor 2); and finally, a three factor model addressing the three constructs separately. The higher order, 1 factor model provided the best fit for the data as will be discussed below in results section. Discriminate validity of the constructs. First, a confirmatory factor analysis (CFA) was conducted to examine the distinctiveness of the leadership and engagement variables. Analysis of moment structure (AMOS; Arbuckle, 1997) with maximum likelihood estimation was used to conduct the CFA. The fit of five nested models ranging from a single-factor model to the hypothesised 7-factor model were examined. The five models were: (a) a one-factor model incorporating all 11 factors; (b) a three-factor model combining the affective leadership variables (factor 1), commitment variables (factor 2), and outcome variables (factor 3); (c) a five-factor model combining the affective leadership behaviours (factor 1), the cognitive commitment variables (factor 2), the affective commitment variables (factor 3), intent to stay (factor 4), and discretionary effort (factor 5); (d) the hypothesised model; and (e) the hypothesised model with cross-loading items removed.

105

The recommended approach to judging the adequacy of a model is to use several fit indices (Dilalla, 2000; Hu & Bentler, 19995; Kline, 1998). A model can be considered to have adequate fit if most or all fit indices are acceptable. The adequacy of models was assessed by the following indices: χ2, χ2/df, Tucker-Lewis Index (TLI: Bentler & Bonett, 1980), Comparative Fit Index (CFI: Bentler, 1990), and the Root Mean Square Error of Approximation (RMSEA; Browne & Cudeck, 1993).The chi-square test examines the differences between the obtained covariance matrix and the predicted covariance matrix. A significant chi-square indicates that the predicted data are different from the obtained data and that the model should be rejected. A limitation of the chi-square test is that it is very sensitive to sample size and rejects almost all null models with the large sample sizes required for SEM. However, it is a useful statistic in comparing nested models (James, Mulahi & Pratt, 1982). The sensitivity of the chi-square to sample size can be reduced by dividing it by the degrees of freedom (χ2/df). Values for the TLI, CFI and GFI can range from zero to 1.00, with values close to 1.00 indicative of good fit. Scores of >.90 are considered representative of a good-fitting model (Hu & Bentler, 1995; Jaccard & Wan, 1996; Kline, 1998). Finally, an RMSEA value of .08 or less is indicative of adequate fit (Dilalla, 2000; Jaccard & Wan, 1996). Chen et al., (2008) more recently proposed two tests that make use of the CI in a consistent way, using the criteria of lower bound of CI≤0:05 and upper bound of CI≤0:1 as two candidate cut-off values. Rejection of the model thus suggests a poor fit in both tests. The results of the analysis are shown in Table 5. Based on sequential chisquare difference tests (James et al., 1982), each more differentiated model fitted the data better. Further, the hypothesised seven-factor model was the only model that satisfied all the fit indices. Some items had high correlated error terms indicating

106

cross-loadings. To ensure the unidimensionality of measurement, some items were removed (Anderson & Gerbing, 1988). Table 3 details the single item removed from each of the scales Table 3: Items removed

Scale

Item

Affective commitment

item 5

Supervisor commitment

item 7

Discretionary effort

item 1

Job feedback C

item 2 (LOS factor)

LOS

item 5 (Goal Clarity factor)

Enabling

item 5

Enabling

item 6

The standardised path coefficients for each of the items in the seven factor model are provided in Table 7. All the items loaded reliably on their predicted factor. The lowest loading was .48. This, in addition to the high internal consistency alphas and moderate intercorrelations among the latent variables provided evidence of the discriminant validity of the measurement scales used.

Table 4: Confirmatory Factor Item Loadings

Factor and Item

LEADER BEHAVIOURS

Loading

107 Showing Genuine Concern Sustains my efforts by demonstrating a genuine interest in me and what I do

.91

Uses knowledge and understanding of what motivates me to achieve goals

.89

Is active in developing my strengths

.93

Is active in supporting my development through coaching /mentoring

.91

Takes time to find out how I feel about being and working in the

.89

organisation/department Is sensitive to my needs/aspirations

.89

Enabling Allows me to lead when the situation requires

.89

Empowers me by enabling me to use discretion in how I perform my job

.92

Delegates effectively because s/he has knowledge of my competence or potential

.92

Empowers me by trusting me take decisions/initiatives on important matters

.92

Being Honest and Consistent Regards the good of the organisation as more important than satisfying personal

.81

ambition Is consistent in his/her behaviour rather than moody or unpredictable

.83

Is honest and open in the way s/he behaves.

.93

Is consistent in what s/he says and what s/he does.

.89

Being Open and Accessible Is committed to developing her/his competence as a leader

.85

Is accessible to staff at different levels

.80

Is prepared to modify decisions etc, rather than being rigidly rule bound

.81

Uses face to face, rather than indirect communication, as and when appropriate

.82

Is approachable, rather than intimidating or status conscious

.85

COGNITIVE ENGAGEMENT Goal Clarity (indicate the degree of certainty you feel about..) The expected results of my work

.85

How my work overall relates to the overall objectives of my work unit

.85

The goals and objectives for my job

.87

Indicate the degree of certainty you feel about: My duties and responsibilities

.90

Job Feedback The feeling I know whether I am performing my job well or poorly.

.84

The opportunity to find out how well I am doing in my job.

.84

The feedback from my supervisor on how well I am doing.

.89

108 To what extent do you receive feedback from your supervisor on your job

.89

performance. To what extent do you find out how well you are doing on the job as you are

.81

working?

Connection I understand the impact of organizational decisions on my job.

.67

I understand exactly how my job contributes to the goals and objectives of my

.87

organization.

AFFECTIVE ENGAGEMENT Affective Commitment I would be very happy to spend the rest of my career in this organisation.

.63

I really feel as if this organization's problems are my own

.56

I do not feel like "part of the family" at my organisation. (R)

.77

I do not feel "emotionally attached" to this organisation (R)

.80

I do not feel a strong sense of belonging to my organisation.(R)

.88

Supervisory Commitment When someone criticises my supervisor, it feels like a personal insult.

.76

When I talk about my supervisor, I usually say "we" rather than "they".

.72

My supervisor's successes are my successes.

.85

When someone praises my supervisor, it feels like a personal compliment.

.89

I feel a sense of "ownership" for my supervisor.

.89

If the values of my supervisor were different, I would not be as attached to my

.70

supervisor.

BEHAVIOURAL ENGAGEMENT Intent to Stay How likely is it that you will actively look for a new job in the next year?

.92

I often think about quitting

.58

I will probably look for a new job in the next year.

.99

Discretionary Effort takes ones job seriously and rarely makes mistakes

.48

Does not mind taking on new or challenging assignments

.76

Tries to self study to increase the quality of work outputs

.64

Often arrives early and starts to work immediately

.48

HIGHER ORDER LOADINGS

109 Affective Leader Behaviours Honest and Consistent

.92

Enabling

.83

Showing Geninue Concern

.91

Being Accessible

.95

Cognitive Engagement Goal Clarity

.82

Job Feedback

.58

Line of Sight

.88

Note: N=635, all loadings are standardised and some items are reverse scored as noted by (R).

Descriptive Statistics and Correlations. The means, standard deviations, intercorrelations and Cronbach alphas for all measures are depicted in Table 6. The correlations were moderate in size. All of the hypothesised relationships were in the predicted direction. The pattern of correlations provided preliminary support for the hypothesised relationships depicted in Figure 6. The internal consistency alphas for all scales were above an acceptable .85 except for the Line of Sight measure (.74) and the discretionary effort measure (.65). Both of which will be discussed in Chapter IV.

110

Table 5: Confirmatory Factor Analyses of Model Fit Indicies χ2

df

χ2/ df

TLI

CFI

RMSEA

15201.03

1378

11.03

.53

.55

.126

Three factor

12471.15

1375

9.07

.62

.64

.113

2729.88*

Five factor

10501.33

1368

7.68

.69

.70

.103

1969.82*

Hypothesised model

4603.94

1355

3.40

.89

.89

.061

5897.39*

Hypothesised model with cross-loading items removed.

3252.56

1058

3.07

.91

.92

.057

1351.38*

Measure

χ2diff

One factor

Note. N=635. The five models were: (a) a one-factor model incorporating all 11 factors; (b) a three-factor model combining the affective leadership variables (factor 1), commitment variables (factor 2), and outcome variables (factor 3); (c) a five-factor model combining the affective leadership behaviours (factor 1), the cognitive commitment variables (factor 2), the affective commitment variables (factor 3), intent to stay (factor 4), and discretionary effort (factor 5); (d) the hypothesised model; (e) the hypothesised model with cross-loading items removed. χ2=difference between observed and obtained covariance matrix; χ2diff = the difference in χ2 from the previous (one factor) model; TLI = Tucker-Lewis index; CFI = comparative fit index; GFI = goodness of fit index; RMSEA = root mean square error of approximation. * p < .001

111

Table 6: Means, Standard Deviations (SD), Inter-correlations and Internal Consistency Alphas for the Variables Mean

SD

1

2

3

4

5

6

7

8

9

10

11

12

Transformational Leadership Affective behaviours

.59

.17

.11

.44

92**

.96)

2. Enabling

.72

.33

88**

73**

.95)

3. Being honest and consistent

.58

.31

91**

79**

71**

.92)

4. Being accessible

.97

.09

92**

80**

74**

83**

.92)

1. shows genuine concern

Attitudinal 5.

Affective commitment

.65

.15

35**

35**

32**

30**

29**

.85)

6.

Supervisor commitment

.34

.14

68**

65**

57**

64**

59**

43**

.91)

7.

Cognitive engagement

.60

76

52**

52**

49**

43**

43**

50**

43**

8.

Goal clarity

.57

.07

41**

39**

42**

34**

33**

41**

30**

88**

.92)

9.

Job feedback

.79

61

63**

70**

52**

52**

53**

35**

52**

64**

41**

.93)

10. Line of sight

.45

.08

34**

33**

32**

29**

28**

45**

31**

87**

64**

37**

.74)

11. Intent to stay

.54

.49

.38**

.37**

.33**

.33**

.33**

.50**

.35**

.38**

.33**

.34**

0.30**

.86)

12. Discretionary effort

.93

66

04

03

05

04

02

15**

11**

16**

08

08*

.22**

.11**

Behavioural Intentions

Note. N = 635. Internal consistency alphas are in parentheses along the diagonal. * p < .05, ** p < .001, two-tailed.

.65)

112

Structural Model Step two of the analysis involved testing the hypothesised structural model. In line with Anderson and Gerbing’s (1988) and Kelloway’s (1995) recommendations for mediation models, I compared two competing models, a fully mediated model and a partially mediated model. The hypothesised fully mediated model is depicted in Figure 8. The hypothesised model had good overall fit with the obtained covariance matrix (χ2=3492.15, df=1068, p< .001, χ2/df =3.27, TLI=.910, CFI=.910, RMSEA=.06). All the paths were significant in the predicted direction. The partially mediated model had additional paths (see Figure 9). First, paths from affective leader behaviours were added to affective and supervisor commitment and intent to stay and discretionary effort. Second, a path from cognitive engagement to intent to stay and discretionary effort was added. Third, a path from supervisor commitment to discretionary effort and intent to stay was added. The partially mediated model had adequate fit with the obtained covariance matrix (χ2 =3308.31, df=1059, p< .001, χ2/df =3.12, TLI=.91, CFI=.92, RMSEA=.06). The addition of these paths led to a significant improvement in fit (χ2diff=183.84, p